The American Journal of Medicine

The American Journal of Medicine Volume 120, Issue 9, Pages.743-824 (September 2007) Commentaries 1. Screening for Prostate Cancer: Current Status ...
Author: Sydney Bates
19 downloads 2 Views 5MB Size
The American Journal of Medicine

Volume 120, Issue 9, Pages.743-824 (September 2007)

Commentaries 1.

Screening for Prostate Cancer: Current Status and Future Prospects Pages 743-745 Nima Sharifi and Barnett S. Kramer

2.

A New Approach to Primary Prevention of Cardiovascular Disease Pages 746-747 John B. Kostis

Reviews 3.

Frailty: An Emerging Geriatric Syndrome Pages 748-753 Nasiya Ahmed, Richard Mandel and Mindy J. Fain

4.

Prerenal Success in Chronic Kidney Disease Pages 754-759 Sheldon Hirsch

Office management: allergy and immunology 5.

Difficult-to-Control Asthma in Adults Pages 760-763 Forest H. Mealey, Nicholas J. Kenyon, Mark V. Avdalovic and Samuel Louie

6.

Evaluation of the Adult with Suspected Immunodeficiency Pages 764-768 Antoine E. Azar and Zuhair K. Ballas

Diagnostic dilemmas 7.

A Vast Differential Pages 769-771 Dana L. Madison

8.

A Rolling Stone Pages 772-774 Marco A. Gonzalez, Joseph J. Sorrento and Ernest Tsao

Images in dermatology 9.

A Pickled Plum of a Problem Pages 775-777 Axel Trautmann, Eva-B. Bröcker and Cornelia S. Seitz

ECG image of the month 10.

Can’t Catch a Breath Pages 778-780 Bishoy Faltas and Ming-Yan Chow

Images in radiology 11.

Narrowing in on a Diagnosis Pages 781-782 Mohit Jain, Satish Gopal and Leslie Trien

Clinical research studies 12.

Efficacy of Short-Course Antibiotic Regimens for Community-Acquired Pneumonia: A Meta-analysis Pages 783-790 Jonathan Z. Li, Lisa G. Winston, Dan H. Moore and Stephen Bent

13.

The Safety of Oral Antifungal Treatments for Superficial Dermatophytosis and Onychomycosis: A Meta-analysis Pages 791-798.e3 Chia-Hsuin Chang, Yinong Young-Xu, Tobias Kurth, John E. Orav and Arnold K. Chan

14.

Impact of Cardiac Rehabilitation on Depression and Its Associated Mortality Pages 799-806 Richard V. Milani and Carl J. Lavie

15.

Independent Vascular and Cognitive Risk Factors for Postoperative Delirium Pages 807-813 James L. Rudolph, Richard N. Jones, Lars S. Rasmussen, Jeffrey H. Silverstein, Sharon K. Inouye and Edward R. Marcantonio

16.

Criteria for the Electrocardiographic Diagnosis of Atrial Flutter Improve Diagnostic Accuracy Pages 814-818 Kenneth M. Weinberg, Pablo Denes, Alan H. Kadish and Jeffrey J. Goldberger

AJM online Clinical research study 17.

Characteristics of Hospitalized Patients with Atrial Fibrillation in Taiwan: A Nationwide Observation Pages 819.e1-819.e7 Cheng-Han Lee, Ping-Yen Liu, Liang-Miin Tsai, Wei-Chuan Tsai, Ming-Tsung Ho, Jyh-Hong Chen and Li-Jen Lin

Brief observations 18.

Medical Errors Arising from Outsourcing Laboratory and Radiology Services Pages 819.e9-819.e11 Brian S. Chasin, Sean P. Elliott and Stephen A. Klotz

19.

Troponin I Levels in Patients with Preeclampsia Pages 819.e13-819.e14 Dominique Joyal, Ferdinand Leya, Megan Koh, Richard Besinger, Ravi Ramana, Steven Kahn, Walter Jeske, Bruce Lewis, Lowell Steen, Ruben Mestril, et al.

Clinical communications to the editor 20.

Rapidly Reversible Cardiogenic Shock as a Pheochromocytoma Presentation Pages e1-e2 Kirsten Brukamp, Simin Goral, Raymond R. Townsend, Frank E. Silvestry and Drew A. Torigian

21.

Cold Water-induced Pulmonary Edema Page e3 Roy Beinart, Shlomi Matetzky, Tova Arad and Hanoch Hod

22.

Unexplained Isolated Elevation in Serum Aspartate Aminotransferase: Think Macro! Pages e5-e6 David A. Sass, Rajagopal Chadalavada and Mohamed A. Virji

23.

Cinacalcet-Associated Graft Dysfunction and Nephrocalcinosis in a Kidney Transplant Recipient Pages e7-e9 Luon W. Peng, Joy L. Logan, Sam H. James, Katherine M. Scott and Yeong-Hau Howard Lien

24.

Q Fever in US Military Returning from Iraq Pages e11-e12 Todd D. Gleeson, Catherine F. Decker, Mark D. Johnson, Joshua D. Hartzell and John R. Mascola

Letters 25.

Treatment of Depression in Patients with Coronary Heart Disease Page e13 Hajime Kojima

26.

The Reply: Page e15 W. Victor R. Vieweg

27.

Spinal Cord Stimulation: A New Form of Pain Modulatory Treatment in Cardiac Syndrome X Page e17 Gregory Angelo Sgueglia and Alfonso Sestito

28.

Opioid Agreement or Patient-Centric Action Plan? Page e19 Jeffrey A. Gudin

29.

Opioid-induced Androgen Deficiency Discussion in Opioid Contracts Page e21 Harry W. Daniell

APM perspectives 30.

Clinical Simulation: Importance to the Internal Medicine Educational Mission Pages 820-824 Paul E. Ogden, Lauren S. Cobbs, Martha R. Howell, Stephen J.B. Sibbitt and Donald J. DiPette

The American Journal of Medicine (2007) 120, 743-745

COMMENTARY

Screening for Prostate Cancer: Current Status and Future Prospects Prostate cancer incurs a huge burden on men, their loved ones, and on the health care system. The American Cancer Society estimates that there will be 218,890 new cases of prostate cancer and 27,050 deaths from prostate cancer in 2007.1 Ironically, the strategy most publicly touted as the most important weapon against prostate cancer—screening—actually may increase the burden of the disease. Yet, the impact on mortality remains unproven despite widespread screening for more than 15 years. In 2003, it was estimated that 1 in 6 men will be diagnosed with prostate cancer, and 1 in 34 will die from prostate cancer.2 Accordingly, given the low mortality rate, less than 1 in 5 men diagnosed with prostate cancer die of prostate cancer. Screening can tap into a very large reservoir of clinically silent prostate cancers. The minimum size of this reservoir was demonstrated in the placebo arm of the randomized Prostate Cancer Prevention Trial (PCPT) of the 5-alpha reductase inhibitor finasteride, in which all study participants were offered “blind” prostate biopsies.3 Over 24% of biopsied men in the placebo arm were found to have prostate cancer. Therefore, less than 1 in 8 men who have prostate cancer die of prostate cancer. Autopsy studies suggest that the proportion could even be considerably lower. This leaves a tremendous gap between the presence of prostate cancer and death resulting from prostate cancer, and for every case of prostate cancer that would not have led to death or morbidity, any therapy administered is ineffective, unnecessary, and usually harmful. One of our goals should be to maximize effectively administered therapies and minimize unnecessary therapies, particularly given that the morbidity associated with surgery and radiation for the treatment of localized prostate cancer often has major consequences on quality of life.4 Although major advocacy and medical organizations, such as the American Cancer Society and the American Urological Association, recommend screening men for prostate cancer with annual PSA (prostate-specific antigen) test and digital rectal examination starting at the age of 50 years, and earlier screening for men at high risk, there is no evidence from prospective randomized trials using an intent-to-treat analysis that screening reduces prostate cancer 0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2006.12.014

mortality. Given the gap between the presence of prostate cancer and prostate cancer mortality, it is unclear how much increased adoption and adherence to screening recommendations is increasing effectively administered therapies and how much it increases unnecessary therapies.5 There are at least 4 major problems in screening for prostate cancer. Problem #1 is the lack of sensitivity and specificity of PSA for the presence of the lethal variety of prostate cancer at a sufficiently early stage. This was one of the results borne out by investigators in the PCPT. It was observed that there is no cutoff value at which PSA has a high sensitivity and high specificity.6 Problem #2 is the uncertainty of the natural history and possibility of death or morbidity for a given case of prostate cancer. The large gap between prostate cancer incidence and mortality has been largely driven for the past 15 years by the national enthusiasm for screening, and this gap is dominated by prostate cancers whose natural history is not defined. Absent results from a definitive randomized controlled trial of screening, we simply cannot know the balance of benefits and harms in such cases. Problem #3 is the uneasiness that comes with not treating a diagnosis of “cancer.” One argument for screening is that aggressive prostate cancer with high Gleason-grade can be treated, whereas low Gleason-grade tumors with a low mortality7 can be monitored without immediate treatment. However, this leaves the patient and clinician with the potentially uneasy and anxiety-provoking task of not treating “cancer.”8 Anxiety can drive overtreatment without benefit. Problem #4 is the lack of gold standard evidence for benefit in morbidity and mortality with prostate cancer screening. A prospective randomized trial is the only direct way to accurately assess the balance of potential benefits and harms incurred by screening. The analytical framework of the US Preventive Services Task Force is useful in avoiding “mental shortcuts” or premature conclusions about screening.9 Steps in the framework include screening the at-risk population, early detection, treatment, adverse effects associated with screening and treatment, intermediate outcomes, and finally, morbidity and mortality. It is important to avoid shortcuts, given that the stakes are so high. It

744

The American Journal of Medicine, Vol 120, No 9, September 2007

has been estimated that over 1 million additional cases of prostate cancer have been detected over the past 25 years as a result of prostate cancer screening.10 The wrong decision, based on faulty evidence or assumptions, can incur substantial harm. Strong evidence is particularly important in promoting interventions to asymptomatic men as opposed to those who are already suffering the effects of a known cancer. Given the high stakes, how can we approach the problems associated with screening for prostate cancer? Problem #1: We can work on preserving the potential benefits of screening while diminishing the harm. For example, we can improve upon sensitivity and specificity for detection of disease where detection makes a difference. There are at least 2 strategies to improve upon the lack of sensitivity and specificity associated with PSA. The first is to change how PSA is used. However, attempts to “improve” upon PSA by lowering the PSA threshold, using age-specific cut-offs, following PSA velocity, measuring free:total PSA ratios, or determining PSA density, have not been reproducible—perhaps because the nonspecificity of PSA for cancer is an underlying structural impediment to any fine tuning. A better strategy may be the development of other biomarkers or biomarker combinations in serum. Recently, a signature of autoantibodies was shown to have a better sensitivity and specificity for prostate cancer compared with PSA.11 Further validation of this technique in prospective, multi-institutional trials will help determine if using this methodology for prostate cancer screening is feasible. Problem #2 is an issue of obtaining better prognosticators for local prostate cancer. Gleason grade is used to distinguish high-risk tumors from more indolent prostate cancer, but further stratification with better certainty for risk of progression is required to determine when local treatment is necessary and when it should be deferred. Preliminary studies suggest that DNA microarray analysis may be predictive of the clinical course and mortality for several cancers, including prostate cancer.12 However, these analyses are generally done on prostatectomy specimens, and using limited tissues from prostate biopsies would be more challenging. Problem #3 in the context of discovering low-grade localized prostate cancer in some clinical settings is using the dreaded word “cancer” without offering immediate treatment. A valuable lesson can be learned from investigators in the ProtecT trial, in which men with a new diagnosis of prostate cancer are offered randomization between radical prostatectomy, radiation therapy, and no immediate treatment. Initially, only about 30% of men in some centers agreed to undergo randomization to “treatment” or “no treatment.” However, equipoise was emphasized in counseling patients and later about 80% of men agreed to be randomized to surgery, radiation, or “active monitoring” (Hamdy F, personal communication). These results confirm that subtle changes in the approach to counseling men with local prostate cancer make important differences.13 Lan-

guage can corrupt thought, and life-changing decisions should not be driven by misunderstanding of the words we use in counseling patients. Problem #4 centers on the elusive evidence that routine prostate cancer screening incurs more benefit than harm. At this point, the benefits are theoretical, the harms inescapable. A recent Scandinavian study of men randomized to radical prostatectomy versus watchful waiting showed a small but statistically significant absolute risk reduction in mortality.14 However, it is imperative to note that prostate cancer screening is not common in this population and only 5% of the patients in this study were diagnosed by screening. Therefore, the mortality benefit seen in this population, in which the majority of men have palpable carcinoma, will not necessarily be germane to a screened population in which the majority of men do not have a palpable abnormality. Screening introduces selection bias, lead-time bias, length bias, and overdiagnosis.15 The Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial randomized more than 75,000 men to a prostate cancer screening arm versus a control arm from 1993 through 2001.16 No mortality benefit has yet been reported, so it will be important to reserve judgment until its definitive endpoints are available. In the meantime, the principle of primum non nocere is important to the assessment of prostate screening and to informed decision-making with patients. Pending results of ongoing randomized prostate cancer screening trials, we should make our uncertainties explicit when discussing screening options with men. Nima Sharifi, MD Medical Oncology Branch National Cancer Institute Bethesda, Md

Barnett S. Kramer, MD, MPH Office of Disease Prevention National Institutes of Health Bethesda, Md

ACKNOWLEDGMENT This research was supported in part by the Intramural Research Program of the NIH, National Cancer Institute. The opinions expressed in this article are those of the authors and do not necessarily represent official opinions or positions of the US federal government of the Department of Health and Human Services.

References 1. American Cancer Society. Overview: prostate cancer. Available at: http://www.cancer.org/docroot/CRI/content/CRI_2_2_1X_How_many_ men_get_prostate_cancer_36.asp?sitearea⫽. Accessed June 7, 2007. 2. Cancer facts & figures, 2003. Atlanta: American Cancer Society; 2003. 3. Thompson IM, Goodman PJ, Tangen CM, et al. The influence of finasteride on the development of prostate cancer. N Engl J Med. 2003;349:215-224. 4. Quek ML, Penson DF. Quality of life in patients with localized prostate cancer. Urol Oncol. 2005;23:208-215.

Sharifi and Kramer

Screening for Prostate Cancer

5. Brawley OW, Kramer BS. Cancer screening in theory and in practice. J Clin Oncol. 2005;23:293-300. 6. Thompson IM, Ankerst DP, Chi C, et al. Operating characteristics of prostate-specific antigen in men with an initial PSA level of 3.0 ng/ml or lower. JAMA. 2005;294:66-70. 7. Albertsen PC, Hanley JA, Fine J. 20-year outcomes following conservative management of clinically localized prostate cancer. JAMA. 2005;293:2095-2101. 8. Tannock IF. Eradication of a disease: how we cured symptomless prostate cancer. Lancet. 2002;359:1341-1342. 9. Harris RP, Helfand M, Woolf SH, et al. Current methods of the US Preventive Services Task Force: a review of the process. Am J Prev Med. 2001;20(3 Suppl):21-35. 10. Welch HG. Should I Be Tested for Cancer: Maybe Not and Here’s Why. Berkley, CA: University of California Press; 2004. 11. Wang X, Yu J, Sreekumar A, et al. Autoantibody signatures in prostate cancer. N Engl J Med. 2005;353:1224-1235.

745 12. Dhanasekaran SM, Barrette TR, Ghosh D, et al. Delineation of prognostic biomarkers in prostate cancer. Nature. 2001;412:822-826. 13. Donovan J, Mills N, Smith M, et al. Quality improvement report: improving design and conduct of randomised trials by embedding them in qualitative research: ProtecT (prostate testing for cancer and treatment) study. Commentary: presenting unbiased information to patients can be difficult. BMJ. 2002;325:766-770. 14. Bill-Axelson A, Holmberg L, Ruutu M, et al. Radical prostatectomy versus watchful waiting in early prostate cancer. N Engl J Med. 2005;352:1977-1984. 15. Kramer BS, Brown ML, Prorok PC, Potosky AL, Gohagan JK. Prostate cancer screening: what we know and what we need to know. Ann Intern Med. 1993;119:914-923. 16. Andriole GL, Levin DL, Crawford ED, et al. Prostate Cancer Screening in the Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial: findings from the initial screening round of a randomized trial. J Natl Cancer Inst. 2005;97:433-438.

The American Journal of Medicine (2007) 120, 746-747

COMMENTARY

A New Approach to Primary Prevention of Cardiovascular Disease Cardiovascular disease remains the most important cause of death and disability in the United States and is expected to achieve this distinction in worldwide statistics in the next few decades.1,2 Although significant advances have been made, the exact basic mechanisms responsible for coronary heart disease, stroke, and heart failure have not been clearly elucidated. However, interventions based on current knowledge such as antihypertensive and lipid-lowering pharmacologic therapy have been shown to provide significant clinical benefits and are incorporated into national guidelines. Existing guidelines for primary prevention have not achieved their objective of controlling risk factors. In the United States, only 36.8% of patients with hypertension have their blood pressure controlled to current guidelines,3 only a minority of patients have controlled low-density lipoprotein (LDL), and a low percentage (⬍10% in 1 study) have both their blood pressure and cholesterol controlled.4 Limitations of current guidelines include their length and complexity and the focus on only 1 risk factor, such as hypertension or dyslipidemia. However, these conditions frequently coexist in the same patient, and for any given level of 1 risk factor, the cardiovascular risk varies widely depending on the presence or absence and the severity of additional risk factors. In addition, most cardiovascular events occur among individuals who have mild or moderate elevations of more than 1 risk factor.5 A better approach is needed.6 The development of 1 set of guidelines for cardiovascular risk reduction in primary prevention based on “global risk,” that is, the risk imposed by the combination of all major risk factors in each individual rather than the current “silo” approaches, is needed.6 Currently, a point score system to calculate 10-year global risk based on the Framingham study is used and included in the National Cholesterol Education Program Adult Treatment Panel III guidelines.7 In these guidelines, 10-year risk thresholds are used to make recommendations for cholesterol-lowering therapy. This approach is flawed even though it is based on global risk. Because age and sex are the strongest determinants of global risk, this method results in not treating or undertreating younger individuals, especially women, even in cases of marked elevations of blood pressure or cholesterol. For 0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2007.02.019

example, the 10-year risk of a 49-year-old woman who does not smoke is less than 10% (less than “moderately high”) even in the presence of high systolic blood pressure (eg, 190 mm Hg) and total cholesterol (eg, 300 mg/dL) in association with low high-density lipoprotein cholesterol (eg, 25 mg/ dL). Another 49-year-old nonsmoking woman with an LDL of 185 mg/dL and no other risk factors does not qualify for LDL-lowering drug therapy. In contrast, nearly all 70-yearold men (even those with ideal levels of the modifiable risk factors) would require treatment for LDL cholesterol more than 130 mg/dL (⬎100 mg/dL as a therapeutic option) because their 10-year risk is “moderately high” (⬎10%). It is more appropriate to base decisions on lifetime global risk. This approach would level the playing field and treat older and younger individuals of both sexes according to the degree of the modifiable risk factors rather than demographics. Treating younger individuals with the intent of preventing atherosclerotic disease rather than focusing on events will probably accrue much larger benefits than current approaches.8 Individuals with non-sense mutations in the gene encoding PCSK-9 have a modest (28%) reduction in LDL cholesterol and a marked (88%) reduction in the risk of coronary heart disease.9 This has been attributed to the lifelong low cholesterol levels and supports the opinion that what matters is not only how low the LDL cholesterol is decreased but also for how long.10 Concerns about treating young individuals with medications for long periods of time (decades) are the expense and safety of long-term therapy, especially during the childbearing years, and the lack of clinical trial evidence of a benefit. The issue of long-term safety is real but exaggerated because both statins and antihypertensive agents have been prescribed to many patients for up to 20 years. If a new serious adverse event materializes after longer use, the administration of the culprit medication(s) can be discontinued. Also, the medications can be discontinued if advances in basic science result in new precisely targeted “silver bullet” agents for prevention. Existing guidelines are founded on “evidence-based medicine” and the Hippocratic principle of “first, do not harm” (“␻␾⑀␭␣␫⑀␫␯ ␩ ␮␩ ␤␭␣␲␶⑀␫␯,” “primum non nocere”), included in Epidemics.11 Interventions are recommended only when a benefit is

Kostis

Primary Prevention of CVD

proven by clinical trials. This approach is ethically asymmetric by placing more emphasis on harms that we may cause than on those we fail to prevent. It is at odds with simple utilitarianism in which moral obligations depend on expected consequences.12 Requiring more justification for an action than for inaction is understandable in situations when the benefits of a given act are small and largely hypothetical, as was the case in the time of Hippocrates. It is not appropriate in the case of quantitatively known risks and benefits of treatment, but also of no treatment, as is the case of lipid-lowering and antihypertensive therapy. In clinical trials (which are usually of short duration), the number needed to treat to prevent 1 event, and the expense of treatment, is higher among young individuals. In a world of limited resources, this may imply that treatment of older individuals (higher global risk) is more cost-effective. On the other hand, ethical principles that may be used to decide how to distribute limited resources in the presence of high demand may favor treating younger individuals. For example, treating the young will result in saving most qualityadjusted life years8 and would provide the opportunity to more individuals to live through all stages of life.13 However, in my opinion, there is no need for “rationing” of antihypertensive and lipid-lowering medications because generic compounds can be provided at a low cost (a few tens or hundreds of dollars per quality-adjusted life year).14 The issue is not one of limited resources but one of application of current knowledge. Simple guidelines clearly communicated and empowerment of both physicians and patients will lead to more effective cardiovascular disease prevention. A new approach to primary prevention could be based on global lifetime risk and implemented with simple guidelines aiming at the prevention of atherosclerotic disease rather than focusing on events and individual risk factors.15 In a fashion similar to traffic laws in which speed limits apply to all (men, women, young, old), I propose 2 numbers (limits) for primary prevention: less than 100 for LDL cholesterol and fasting glucose, and less than 130 for systolic blood pressure and non– high-density lipoprotein cholesterol. “No smoking any time” can substitute for “No parking anytime.” These new limits will save many more lives than all traffic rules put together.

747 John B. Kostis, MD UMDNJ-Robert Wood Johnson Medical School Division of Cardiology New Brunswick, NJ

References 1. American Heart Association. Heart disease and stroke statistics 2006 update. Available at: http://www.americanheart.org/presenter.jhtml? identifier⫽3018163. Accessed January 29, 2007. 2. Murray CJ, Lopez AD. Alternative projections of mortality and disability by cause 1990-2020: Global Burden of Disease Study. Lancet. 1997;24:1498-1504. 3. Ong KL, Cheung BMY, Man YB, et al. Prevalence, awareness, treatment, and control of hypertension among United States adults. 19992004. Hypertension. 2007;49:1:69-75. 4. O’Meara JG, Kardia SL, Armon JJ, et al. Ethnic and sex differences in the prevalence, treatment, and control of dyslipidemia among hypertensive adults in the GENOA study. Arch Intern Med. 2004; 164:1313-1318. 5. Kannel WB. Risk stratification in hypertension: new insights from the Framingham Study. Am J Hypertens. 2000;13(1 Pt 2):3S-10S. 6. Kostis JB. Treating hypercholesterolemia and hypertension based on lifetime global risk. Am J Cardiol. In press. 7. Expert Panel on Detection, Evaluation and Treatment of High Blood Cholesterol in Adults. Executive summary of the third report of the national cholesterol education program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (adult treatment panel III). JAMA. 2001;285:2486-2497. 8. Jackson PR. The influence of absolute cardiovascular risk, patient utilities and costs on the decision to treat hypertension: a Markov decision analysis. J Hypertension. 2003;21:1631-1634. 9. Cohen JC, Boerwindle E, Mosley TH Jr, Hobbs HH. Sequence variations in PCSK9, low LDL, and protection against coronary heart disease. N Engl J Med. 2006;354:1264-1272. 10. Brown MS, Goldstein JL. Lowering LDL—not only how low, but how long? Science. 2006;311:1721-1723. 11. Hippocrates. Epidemics, Book. I, Sect. V. 12. Baron J. Do no harm. In: Messick DM, Tenbrunsel AE, eds. Codes of Conduct: Behavioral Research into Business Ethics. New York, NY: Russell Sage Foundation; 1996:197-213. 13. Emanuel EJ, Wertheimer A. Who should get influenza vaccine when not all can? Science. 2006;312:854-855. 14. Anderson GF, Chu E. Expanding priorities— confronting chronic disease in countries with low income. N Engl J Med. 2007;356;3: 209-211. 15. Kostis JB. Disputation on the use of age in determining the need for treatment of hypercholesterolemia and hypertension. J Clin Hypertens. 2006;8:519-520.

The American Journal of Medicine (2007) 120, 748-753

REVIEW

Frailty: An Emerging Geriatric Syndrome Nasiya Ahmed, MD,a,b Richard Mandel, MD,a Mindy J. Fain, MDa,b a

Southern Arizona VA Health Care System, Tucson and bUniversity of Arizona, College of Medicine, Tucson ABSTRACT Frailty is a new and emerging syndrome in the field of geriatrics. The study of frailty may provide an explanation for the downward spiral of many elderly patients after an acute illness and hospitalization. The fact that frailty is not present in all elderly persons suggests that it is associated with aging but not an inevitable process of aging and may be prevented or treated. The purpose of this article is to review what is known about frailty, including the definition, epidemiology, and pathophysiology, and to examine potential areas of future research. © 2007 Elsevier Inc. All rights reserved. KEYWORDS: Aging; Frailty; Geriatrics

The relentless path to old age, frailty, and death has been a hopeless enigma of medicine. In Ptolemy’s treatise, the last age of life from 68 years to death was described as “dispirited, weak, easily offended, and hard to please.” Aristotle further explained this stage as a period when the body’s heat dissipates and is no longer able to provide energy and balance; the loss of inward heat depressed the spirit, caused illness, and decreased strength, and eventually human beings lost their passion for life and succumbed to death.1 In 1908, Eli Metchnikoff asked, “How can we transform to a normal and physiological condition, old age, at present utterly pathological, unless we first understand the most intimate details of its mechanism?”2 This question is just beginning to be answered almost a century later as geriatricians struggle to define frailty. This article will review the definition, epidemiology, pathophysiology, and treatment of frailty, and examine potential areas of future research.

FRAILTY DEFINED Frailty, a fairly common biological syndrome in the elderly, is identified by decreased reserves in multiple organ systems. It may be initiated by disease, lack of activity, inadequate nutritional intake, stress, and/or the physiologic changes of aging. Frailty develops slowly in a stepwise Requests for reprints should be addressed to Mindy J. Fain, MD, University of Arizona, College of Medicine, Section of Geriatrics and Gerontology, PO Box 245069, Tucson, AZ 85724-5069. E-mail address: [email protected]

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2006.10.018

process, with increments of decline precipitated by acute events. It is manifested as loss of skeletal muscle mass (sarcopenia), abnormal function in inflammatory and neuroendocrine systems, and poor energy regulation. In the frail elderly, there is homeostenosis, or a decreased ability in the body’s physiologic response to maintain homeostasis in times of acute stress. In essence, frailty is a product of “excess demand imposed upon reduced capacity.”3 Once the elderly become frail, there is often a rapid, progressive, and self-perpetuating downward spiral toward failure to thrive and death.4

BACKGROUND Contrary to popular belief, not all elderly are frail.5 Only 3% to 7% of elderly persons between the ages of 65 to 75 years are frail.6 The incidence of frailty increases with age, reaching more than 32% in those aged more than 90 years.7 Furthermore, once a person is pre-frail, he or she is more likely to progress to frailty, thus emphasizing the downward spiral affect of this syndrome.8 Frailty can be a primary or secondary diagnosis. Notably, 7% of the frail elderly have no illness, and 25% have only 1 comorbid diagnosis.6 Researchers have demonstrated that even when individuals with acute and chronic medical conditions were excluded, 7% of the population aged more than 65 years and 20% of the population aged more than 80 years were frail.5 On the other hand, frailty may occur as a result of an acute event or the end stage of many chronic conditions, including atherosclerosis, infection, malignancy, and

Ahmed et al

Frailty

749

depression.5 For example, severe congestive heart failure aged more than 85 years. The incidence of frailty increases can lead to decreased activity, decreased nutritional intake, with age and will become more prevalent as our population and increased inflammation and circulating cytokines. The continues to grow old.13 Studies show that 3% to 7% of combination of these factors eventually leads to cardiac people aged more than 65 years are frail6; this percentage cachexia and frailty. increases to 20% to 26% for elderly persons in their 80s5,6 and to 32% for those in their 90s.7 The phenotypic picture of frailty The 4-year incidence of frailty in can be confused easily with disabilthe elderly is 7%10; a similar study ity; however, investigators have CLINICAL SIGNIFICANCE in the Hispanic population showed sought to differentiate them. Disthe 7-year incidence to be 7.9%.8 ability, defined as the inability to ● When compared with non-frail elderly, In general, the elderly population perform activities of daily living pre-frail elderly have an increased risk of has a greater percentage of fe(ADL), instrumental activities of falls, institutionalization, and mortality males and frailty is considerably daily living (IADL), or difficulty but not as high as the frail elderly. higher in women.14 After adjustwith mobility, does not affect the ment for age, race, sex, smoking, body across multiple organ sys● During the pre-frail stage, the frailty and comorbid illness, frail patients tems.9 Among frail elderly persons, syndrome may be reversed. only 60% have difficulty complethave 1.2- to 2.5-fold increase in ● Multiple studies have shown the benefit ing IADL, and 27% cannot comtheir risk for falls, decreased moof exercise, stretching, resistance trainplete ADL; furthermore, only 28% bility, worsening ADL, institutioning, and tai chi on frailty. Frailty markers of disabled elderly persons are alization, and death.10 A separate, 10 have been shown to improve after 30 to frail. Even when adjustments are cross-sectional observational study made for disability and comorbidreported that frail patients had a sig60 minutes of exercise, done 3 times a ity, elderly persons with a diagnosis nificantly increased risk of cardioweek. of frailty continue to have a higher vascular disease, hypertension, can● By recognizing the frailty syndrome and mortality rate.11 Although disability cer, and death, even after adjusting suggesting lifestyle changes, physicians may contribute to frailty and vice for chronic conditions.15 As the may help patients prevent comorbidities versa, the 2 diagnoses are distinct elderly population continues to later in life. from each other. grow, the impact of frailty will be Frailty has been described as felt throughout families and perhaving a continuum. The initial vade our economic, health care, stage of frailty, in which patients and social systems. demonstrate fewer than 3 of the characteristics diagnostic of frailty, is referred to as the pre-frail stage. Studies show that A WORKING DIAGNOSIS OF FRAILTY pre-frail elderly persons are more likely than non-frail elThe diagnosis of frailty has taken time to evolve. Initial studies derly persons to develop the full syndrome.10 Pre-frail elshowed a decrease in strength and balance to be predictors of derly persons also have an increased risk of falls, institufrailty.16,17 However, in the elderly, multiple causes can lead to tionalization, and mortality, but not as high as frail elderly such a state, including immobility, decreased appetite, poor persons.10 It is at the pre-frail stage that investigators benutrition, and chronic illness. On this basis, a “cycle of frailty” lieve the frailty syndrome may be reversed. The end stage of was hypothesized. The Figure helps conceptualize this cycle 6 the frailty continuum is failure to thrive. Elderly patients, and highlights the interdependence of various factors that may with or without comorbidity, eventually reach an irreverscause patients to enter this cycle. For example, poor dentition ible stage of functional decline, progressive apathy, and itself cannot cause frailty; however, it can lead to chronic 12 decreased appetite that ultimately ends in death. undernutrition and eventually sarcopenia, thus placing the paA more recent study expands on the concept of a contient in the cycle of frailty. Poor dentition can also lead to tinuum of frailty, defining it as both a static and dynamic infection and periodontal disease, which may cause an increase diagnosis. Dynamic frailty is defined as a decline in the in inflammatory markers and cytokines, once again placing the measurement of frailty markers over a 3-year period, with or patient in this cycle. without a diagnosis of frailty. Dynamic frailty, even when More recently, Fried10 conducted a detailed study anaadjusted for age, education, disability, chronic disease, and lyzing multiple demographic, medical, and laboratory markstatic frailty markers, was still associated with an increased ers to statistically determine a definition of frailty. A diag11 mortality; this increase was greater in women than in men. nosis of frailty requires 3 of the following 5 characteristics (Table):

SIGNIFICANCE Why is frailty important? Currently 20% of the population of the United States are aged more than 65 years, and the most rapidly growing segment of our population are those

● ● ●

Decreased walk time, as defined by a 15-foot walk test. Decreased grip strength, measured by a dynamometer. Decreased physical activity, measured by the Minnesota Leisure Time Activity Questionnaire.18

750

The American Journal of Medicine, Vol 120, No 9, September 2007

Figure

● ●

The frailty cycle.

Exhaustion, measured by the Center for Epidemiologic Studies Depression Scale. More than 10 pounds or 5% of weight loss in the last 1 year.

This set of criteria indirectly measures the manifestations of frailty, most important, sarcopenia as measured by grip strength, malnutrition as measured by weight loss, and fatigue as measured by the Center for Epidemiologic Studies Depression Scale. Recently, these criteria were applied to data from the Women’s Health and Aging Studies. By using this definition, a similar incidence and prevalence of frailty

Table

were observed; in addition, statistical correlations between the incidence of frailty and the development of disability, falls, institutionalizations, and mortality mirrored the results found in Fried’s study.19 Although Fried’s definition of frailty is most widely accepted and will be the definition referenced in this article, other indices of frailty have also been published.20-22 Not all that appears to be frailty is frailty. The differential diagnosis of frailty includes congestive heart failure, polymyalgia rheumatica, Parkinson disease, rheumatoid arthritis, occult malignancy, and infection. A new onset or an

Criteria Used to Define Frailty One Must Have 3 of the Following 5 Criteria to Be Frail Male

Weight Loss 15-Foot Walk Time Grip Strength

Physical Activity (MLTA) Exhaustion

Greater than 10 lbs or 5% of weight loss in the last Height ⱕ173 cm ⱖ7 seconds Height ⬎173 cm ⱖ6 seconds BMI ⱕ24 ⱕ29 BMI 24.1-26 ⱕ30 BMI 26.1-28 ⱕ30 BMI ⬎28 ⱕ32 ⬍383 kcal/wk A score of 2 or 3 on either question on the CES-D*

Female year Height ⱕ159 cm Height ⬎159 cm BMI ⱕ23 BMI 23.1-26 BMI 26.1-29 BMI ⬎29 ⬍270 kcal/wk

ⱖ7 seconds ⱖ6 seconds ⱕ17 ⱕ17.3 ⱕ18 ⱕ21

*How often in the last week did you feel this way? (a) I felt that everything I did was an effort. (b) I could not get going. 0 ⫽ 1 day; 1 ⫽ 1-2 days; 2 ⫽ 3-4 days; 3 ⫽ more than 4 days. BMI ⫽ body mass index; MLTA ⫽ Minnesota Leisure Time Activity Questionnaire; CES-D ⫽ Center for Epidemiologic Studies Depression Scale.

Ahmed et al

Frailty

exacerbation of any of these diseases could present as frailty; however, all of these illnesses are treatable and should be excluded before attributing a decline to frailty.6

Other Markers of Frailty Although not part of the definition, other demographic and neuroendocrine markers also have been associated with frailty. Demographic markers include female sex, African-American race, lower education level, lower income, chronic illness, and disability.10,23 In the previous study, chronic illnesses associated with frailty included cardiovascular disease, pulmonary disease, arthritis, and diabetes; cancer was an exclusion criteria. Cognitive limitations and depression were also more prevalent, despite the exclusion of patients taking antidepressants or having a mini-mental status examination score of less than 18. Researchers also have evaluated the inflammatory and neuroendocrine changes in frailty. Walston and others7 looked extensively at frailty markers in the Cardiovascular Health Study Cohort. When excluding patients with diabetes or cardiovascular disease, frail patients had an increased C-reactive protein (CRP) and fibrinogen (odds ratio 2.03.0). CRP has been shown to activate the inflammation and clotting cascade, including d-dimer and factor VIII, both of which are elevated in frail patients. CRP also causes an increase in circulating interleukin-6; Leng and others24 confirmed that frail patients have higher levels of interleukin-6. Glucose intolerance also was noted; both fasting/postprandial glucose and insulin levels were elevated in frail patients (odds ratio 1.5-2.6).25 Glucose intolerance is often associated with sarcopenia. In addition, recent research shows that hyperinsulinemia and hypertriglyceridemia may be associated with cognitive decline and leptin resistance, resulting in appetite suppression and decreased nutritional intake. Serum levels of insulin-like growth factor-1 and dehydroepiandrosterone were significantly lower in frail patients as well.26 Dehydroepiandrosterone, a weak androgenic steroid that is a precursor of testosterone, plays a role in maintaining muscle mass and suppressing inflammation. Insulin-like growth factor-1 stimulates growth hormone release that regulates cell growth and development, and is often decreased in patients with diabetes or malnutrition. No relationship between frailty and lipid profiles or albumin has been seen.21 It is important to note that it remains unknown whether the above neuroendocrine changes cause frailty or whether the reverse is true.

Markers of Aging Before the formulation of Fried’s definition, researchers using various criteria for frailty noted other demographic trends in patients who appeared to be frail. Not all of these markers have been validated using Fried’s definition. One study linked frailty in men to depression, inactivity, weight loss, decreased peak flow, and decreased cognition. Furthermore, a decline in peak flow, cognition,

751 vision, and physical activity over a 3-year period was linked to frailty in women.11 Other studies have found incontinence, poor hearing, and a feeling of loss of control over one’s life to be associated with frailty.11 A study based in China found that social factors, such as limited contact with relatives, blue-collar occupations, and absence of religious or community activities, were more likely to be seen in the frail elderly.27 Many gerontologic studies have shown inflammatory and neuroendocrine markers linked to aging, but these experiments have not been reproduced in the frail population.28,29 In times of stress, cortisol secretion and pituitary reactivity are increased in older adults.30 Although increased levels of cortisol have not been directly linked to frailty, this can lead to the frailty phenotype, including increased risk of infection, sarcopenia, and insulin resistance. Tumor necrosis factor-␣ has also been linked to aging and sarcopenia; studies attempting to link this cytokine to frailty are currently under way.31 More recently, leptin, a hormone that controls appetite and body fat stores, and ghrelin, a hormone that inhibits leptin, have been shown to be links in the endotoxin-induced cachexia suspected in frailty.32 A recent study, using an alternative definition of frailty, showed that elderly persons had lower 25(OH)D levels (odds ratio 2.6).33

Frailty and Its Causes What causes frailty? Although many studies have shown associations between the incidence of frailty and factors that may lead to the condition, a definitive correlation has not been established. It also remains unclear why factors trigger frailty in some individuals but not in others. Experts speculate that certain environments, medications, age-related changes, and diseases make a particular genotype of people vulnerable to frailty.5,34 Data from the Women’s Health Initiative study showed that obesity, anorexia, smoking, and depression may lead to the development of frailty.35 Multiple aging studies showed that cumulative predictors over 3 decades, including heavy drinking, cigarette smoking, physical inactivity, depression, social isolation, poor perceived health, and chronic illnesses, led to increased morbidity in the elderly but not necessarily frailty.36 At the cellular level, Bortz37 proposed a theory referred to as “the physics of frailty,” which postulates that a loss of functional competence at a cellular level, along with thermodynamic decline and a loss of energy stores, leads to physiologic decline. Walston38 further adds that the cause of frailty might be at the molecular level; mechanisms related to aging, such as oxidative damage, telomere shortening, gene expression changes, and cellular senescence, may contribute to the dysregulated inflammatory and neuroendocrine signaling that leads to frailty. Genetic research also has been done linking Apo E to frailty as well.39

752

The American Journal of Medicine, Vol 120, No 9, September 2007

CLINICAL RESEARCH Research also has begun in the clinical realm. Although a detailed discussion of clinical research is beyond the scope of this article, a general overview is provided. Atherosclerosis contributes to frailty by decreasing blood flow and oxygenation to muscles, leading to sarcopenia. It indirectly contributes to cognitive impairment, through strokes, and to decreased physical activity in congestive heart failure and myocardial infarction.31 The Cardiovascular Health Study showed that cardiovascular disease, most notably heart failure (odds ratio 7.5), was associated with an increased likelihood of frailty.40 In elderly patients with no history of cardiovascular disease, subclinical cardiovascular disease, measured by carotid ultrasound, ankle-brachial index, and left ventricular hypertrophy, was also linked to frailty.40 Obesity also carries a high risk of frailty.41,42 Anemia is increased in frail patients; researchers have noted a direct correlation between decreased hemoglobin and increased frailty.24,43 Furthermore, anemia in the frail population has been linked to elevation of interleukin-6.44 Patients with chronic renal insufficiency, after adjustment for other comorbidities, also have a higher risk of frailty.45 Research also has shown that frailty and concurrent inflammation affect the treatment of infectious diseases.46 Chronic cytomegalovirus infection has been associated with frailty, but further studies are needed to establish a causal relationship.47 Frailty is more often seen in both depressed and cognitively impaired patients.48,49 Increased inflammation, but not necessarily frailty, has been shown to prevent wound healing as well.

Treatment There is limited treatment available for frailty. The goal of initial treatment is the optimal management of all medical illnesses that may cause frailty. The second step is to prevent sarcopenia through muscle-strengthening exercises.50,51 Multiple studies have shown the benefit of exercise, stretching, resistance training, and tai chi on frailty. Frailty markers have been shown to improve after 30 to 60 minutes of exercise, done 3 times per week, for 3 to 6 months. Exercise also has been shown to decrease the level of inflammatory markers. However, studies have not been done to show improvement in mortality with exercise; it is known, however, that exercise does not improve morbidity or mortality in the general elderly population. Nutritional research conducted on the elderly population may shed some light on possible treatments for frailty. It has been found that increased caloric intake has little benefit in improving health status in the elderly; however, there does seem to be some benefits gained from a nutritional program when combined with an exercise program to increase expenditures by 1000 kcal/week.6 The appetite stimulants megestrol acetate and dronabinol have been of minimal benefit in improving appetite in geriatric nursing home patients; moreover, these drugs are associated with significant adverse effects.12

It has been noted that as people age, levels of testosterone and DHEA, 2 hormones that have been shown to maintain muscle strength, decrease. However, investigators have been unable to demonstrate any benefit from hormone replacement.25 Growth hormone replacement also has failed to produce any positive effects in frail elderly persons, although short-term treatment has improved malnutrition and functional status.25 Last, a study done on frail elderly persons showed that a positive attitude prevented both the physical and mental decline of frailty.52

CONCLUSION An understanding of what Eli Metchnikoff once referred to as the “intimate details” of the mechanisms behind an “utterly pathological” old age has just begun. It is now known that old age is not synonymous with the frailty syndrome. Still, frailty has only recently been defined, and frailty research is in its early stages. Acceptance of a formalized set of criteria for the diagnosis of frailty will facilitate studies. Potential areas of investigation include the underlying pathophysiology of primary frailty and the contributions of other comorbid diagnoses to the clinical picture. Further investigation may illuminate its symptoms and impact on the body’s various organ systems, and how best to treat those who are frail. By identifying the risk factors for frailty, a window of prevention before the development of frailty may be found. Last, the effect of frailty on other organ systems may be elucidated.53 Successful investigations may lead to pharmacologic or other treatment modalities to address the needs of frail and pre-frail patients, and prevent both the development and the effects of this geriatric syndrome.

References 1. Cole T. The Journal of Life. New York: Cambridge University Press; 1997. 2. Achenbaum WA. Crossing Frontiers: Gerontology Emerges as a Science. New York: Cambridge University Press; 1995. 3. Powell C. Frailty: help or hindrance. J R Soc Med. 1997;90:23-26. 4. Hamerman D. Toward an understanding of frailty. Ann Intern Med. 1999;130(11):945-948. 5. Wilson JF. Frailty—and its dangerous effects—might be preventable. Ann Intern Med. 2004;141:489-492. 6. Fried LP, et al. Frailty. In: Cassel C, Leipzig R, Cohen H, Larson E, Meier D, eds. Geriatric Medicine: An Evidence-Based Approach. 4th edition. New York: Springer-Verlag; 2003:1067-1074. 7. Walston J, McBurnie MA, Newman A, et al. Frailty and activation of the inflammation and coagulation systems with and without clinical comorbidities: results from the Cardiovascular Health Study. Arch Intern Med. 2002;162(20):2333-2342. 8. Gill T, Gahbauer EA, Allore HG, et al. Transitions between frailty states among community-living older persons. Arch Intern Med. 2006; 166:418-423. 9. Fried LP, Ferucci L, Darer J, et al. Untangling the concepts of disability, frailty, and comorbidity: implications for improved targeting and care. J Gerontol A Biol Sci Med Sci. 2004:59(3):255-263. 10. Fried LP, Tangen CM, Walston J, et al. Frailty in older adults: evidence for a phenotype. J Gerontol. 2001;56A:M146-M156. 11. Puts MTE, Lips P, Deeg DJ. Sex differences in the risk of frailty for mortality independent of disability and chronic diseases. J Am Geriatr Soc. 2005;53:40-47.

Ahmed et al

Frailty

12. Robertson RG. Geriatric failure to thrive. Am Fam Physician. 2004; 70:343-350. 13. Klein BE, Klein R, Knudtson MD, et al. Frailty, morbidity, and survival. Arch Gerontol Geriatr. 2005;41:141-149. 14. Walston J, Fried LP. Frailty and the older man. Med Clin North Am. 1999;83:1173-1194. 15. Klein BE, Klein R, Knudtson MD, Lee KE. Frailty, morbidity and survival. Arch Gerontol Geriatr. 2005;41:141-149. 16. Dayhoff NE, Suhrheinrich J, Wigglesworth J, et al. Balance and muscle strength as predictors of frailty in older adults. J Gerontol Nurs. 1998;24:18-27. 17. Carriere I, Colvez A, Favier F, et al. Hierarchical components of physical frailty predicted incidence of dependency in a cohort of elderly women. J Clin Epidemiol. 2005;58:1180-1187. 18. Taylor HL, Jacobs DR Jr, Schucker B, et al. A questionnaire for the assessment of leisure time physical activities. J Chronic Dis. 1978;31: 741-755. 19. Bandeen-Roche K, Xue QL, Ferucci L, et al. Phenotype of frailty: characterization in the women’s health and aging studies. J Gerontol A Biol Sci Med Sci. 2006;61:260-261. 20. Rockwood K. What would make a definition of frailty successful? Age Ageing. 2005;34:432-434. 21. Jones KM, Song X, Rockwood K. Operationalizing a frailty index from a standardized comprehensive geriatric assessment. J Am Geriatr Soc. 2004;52:1929-1933. 22. Chin P, Dekker JM, Feskens EJ, et al. How to select a frail elderly population: a comparison of three working definitions. J Clin Epidemiol. 1999;52:1015-1021. 23. Hirsch C, Anderson ML, Newman A, et al. The association of race with frailty. Ann Epidemiol. 2006;16:545-553. 24. Leng S, Chaves P, Koenig K, et al. Serum interleukin-6 and hemoglobin as physiological correlates in the geriatric syndrome of frailty: a pilot study. J Am Geriatr Soc. 2002;50:1268-1271. 25. Morley J, Kim MJ, Haren MT. Frailty and hormones. Rev Endocr Metab Disord. 2005;6:101-108. 26. Leng S, Cappola AR, Anderson RE, et al. Serum levels of IGF-1 and DHEA-S and their relationship with IL-6 in the geriatric syndrome of frailty. Aging Clin Exp Res. 2004;16:153-157. 27. Woo J, Goggins W, Sham A, et al. Social determinants of frailty. Gerontology. 2005;51:402-408. 28. Joseph C, Kenny AM, Taxel P, et al. Role of endocrine-immune dysregulation in osteoporosis, sarcopenia, frailty, and fracture risk. Mol Aspects Med. 2005;26:181-201. 29. Vanitallie TB. Frailty in the elderly: contributions of sarcopenia and visceral protein depletion. Metabolism. 2003;52:22-26. 30. Lamberts SW, van den Beld AW, van der Lely AJ. The endocrinology of aging. Science. 1997;278:419-424. 31. Hamerman D. Toward an understanding of frailty. Ann Intern Med. 1999;130:945-948. 32. Dixit VD, Schaffer EM, Pyle RS, et al. Ghrelin inhibits leptin-and activation induced proinflammatory cytokine expression by human monocytes and T-cells. J Clin Invest. 2004;114:57-66. 33. Puts M, Visser M, Twisk JW, et al. Endocrine and inflammatory markers as predictive of frailty. Clin Endocrinol. 2005;63:403. 34. Morly J, Perry HM, Miller DK. Editorial: something about frailty. J Gerontol A Biol Sci Med Sci. 2002;57:M698-704.

753 35. Woods NF, LaCroix AZ, Gray SL, et al. Frailty: emergence and consequences in women aged 65 and older in the Women’s Health Initiative Observational Study. J Am Geriatr Soc. 2005;53:1321-1330. 36. Strawbridge WJ, Shema SJ, Balfour JL, et al. Antecedents of frailty over three decades in an older cohort. J Gerontol A Biol Sci Med Sci. 1998;53B:9-16. 37. Bortz WM. The physics of frailty. J Am Geriatr Soc. 1993;41:10041008. 38. Walston J. Frailty—the search for underlying causes. Sci Aging Knowledge Environ. 2004;2004:pe4. 39. Gerdes LU, Jeune B, Ranberg KA, et al. Estimation of Apo E genotype-specific relative mortality risks from the distribution of genotypes in centenarians and middle-aged men: Apo E gene is a frailty gene, not a longevity gene. Genet Epidemiol. 2000;19:202-210. 40. Newman AB, Gottdiener JS, Mcburnie MA, et al. Associations of subclinical cardiovascular disease with frailty. J Gerontol A Biol Sci Med Sci. 2001;56:M158-166. 41. Villareal DT, Banks M, Sinacore DR, et al. Effect of weight loss and exercise on frailty in obese older adults. Arch Intern Med. 2006;166: 860-866. 42. Blaum CS, Xue QL, Michelon E, et al. The association between obesity and the frailty syndrome in older women: the Women’s Health and Aging Studies. J Am Geriatr Soc. 2005;53:927-934. 43. Chaves PH, Semba RD, Leng SX, et al. Impact of anemia and cardiovascular disease on frailty status of community-dwelling older women: the Women’s Health and Aging Studies I and II. J Gerontol A Biol Sci Med Sci. 2005;60:729-735. 44. Ershler WB. Biological interactions of aging and anemia: a focus on cytokines. J Am Geriatr Soc. 2003;51:S18-S21. 45. Shlipak MG, Stehman-Breen C, Fried LF, et al. The presence of frailty in elderly persons with chronic renal insufficiency. Am J Kidney Dis. 2004;43:861-867. 46. High KP, Bradley S, Loeb M, et al. A new paradigm for clinical investigation of infectious syndromes in older adults: assessment of functional status as a risk factor and outcome measure. Clin Infect Dis. 2005;40:114-122. Epub 2004 Dec 6. 47. Schmalts HN, Fried LP, Xue QL, et al. Chronic cytomegalovirus infection and inflammation are associated with prevalent frailty in community dwelling older women. J Am Geriatr Soc. 2005;53:747754. 48. Fried LP. Frailty in older adults: evidence for a phenotype. J Gerontol. 2001;56A:M146-156. 49. Katz IR. Depression and frailty: the need for multidisciplinary research. Am J Geriatr Psychiatry. 2004;12:1-6. 50. Binder EF, Schechtman KB, Ehsani AA, et al. Effects of exercise training on frailty in community-dwelling older adults. J Am Geriatr Soc. 2002;50:1921-1928. 51. Wolf SL, Barnhart HX, Kutner NG, et al. Reducing frailty and falls in older persons: an investigation of tai chi and computerized balance training. J Am Geriatr Soc. 2003;51:1794-1803. 52. Ostir GV, Ottenbacher KJ, Markides KS. Onset of frailty in older adults and the protective role of positive affect. Psychol Aging. 2004; 19:402-408. 53. Walston J, Hadley EC, Ferrucci L, et al. Research agenda for frailty in older adults: toward a better understanding of physiology and etiology: summary from the American Geriatrics Society National Institute on Aging Research Conference on Frailty in Older Adults. J Am Geriatr Soc. 2006;54:991-1001.

The American Journal of Medicine (2007) 120, 754-759

REVIEW

Prerenal Success in Chronic Kidney Disease Sheldon Hirsch, MD Lakeside Nephrology, Chicago, Ill. ABSTRACT Renin-angiotensin system inhibitors and diuretics are commonly prescribed to patients with chronic kidney disease to reduce systemic blood pressure. The renin-angiotensin inhibitors also reduce intraglomerular pressure. The lower pressures may result in initial increases in the serum creatinine. The long-term renoprotection provided by these therapies establishes the basis for tolerating the initial increases. However, physicians are sometimes reluctant to continue these treatments when the serum creatinine increases. Several reasons for this reluctance are discussed, including the failure to distinguish between hemodynamic- and parenchymal-mediated changes in kidney function. In addition, the lack of a formal term and place in our diagnostic algorithm for increases in serum creatinine that derive from ultimately beneficial hemodynamic alterations may be a hindrance. The term “prerenal success” is proposed to describe hemodynamic alterations associated with improved prognosis and is placed in a new algorithm. Finally, recent literature describing harmful effects of increases in serum creatinine in other cohorts is reviewed; these cohorts are sufficiently different from the stable chronic kidney disease patient that the results ought not to be extrapolated. © 2007 Elsevier Inc. All rights reserved. KEYWORDS: Chronic kidney disease; Prerenal success; Renin-angiotensin system inhibitors

The progression of chronic kidney disease to end-stage renal disease may be decelerated by blood pressure control and reduction of proteinuria and intraglomerular pressure with renin-angiotensin system inhibitors (RASIs).1-6 However, the reduction of systemic and intraglomerular capillary pressures may result in an initial increase in the serum creatinine,2,7,8 particularly in patients with chronic kidney disease who have impaired renal autoregulation.9 In the interests of long-term renal preservation, tolerance of an initial hemodynamic-mediated increase in serum creatinine is recommended.9-12 This sequence of events is not limited to RASI therapy. Lowering blood pressure with antihypertensive medicines other than RASIs also may result in an initial hemodynamicmediated increase in the serum creatinine.7,8 Continuing non-RASI antihypertensive medicines even after the initial increase has provided long-term renoprotection, as blood pressure control per se is renoprotective.7,13,14 Blood pressure reduction to levels associated with optimal long-term prognosis in chronic kidney disease often Requests for reprints should be addressed to Sheldon Hirsch, MD, Lakeside Nephrology, 55 East Washington St., Chicago, IL 60602. E-mail address: [email protected]

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2007.02.025

requires diuresis, as chronic kidney disease is typically a volume expanded state.15,16 Although specific renoprotection apart from an antihypertensive effect is not ascribed to diuretics (as it is for RASIs), in several studies that demonstrated long-term renoprotection of RASIs in chronic kidney disease, ⬃80% or more of patients also received diuretics for further blood pressure control.1,4,6 As blood pressure control in patients with chronic kidney disease is required to forestall progressive disease and often requires diuretics, the latter are prominently recommended by major advisories for co-administration with RASIs.12,17,18 These recommendations, however, are not always followed in clinical practice; in particular, when the serum creatinine increases, physicians may diagnose acute renal failure, or “acute on chronic” renal failure, and discontinue RASIs or diuretics. The reluctance to tolerate increases in serum creatinine may variably derive from an incomplete understanding of the complexity of the determinants of glomerular ultrafiltration, limitations of our current approach to elevations in serum creatinine, and unjustified extrapolation from data on increases in serum creatinine in other contexts to the treatment of chronic kidney disease. This article reviews these issues.

Hirsch

Prerenal Success in Chronic Kidney Disease

THE DETERMINANTS OF GLOMERULAR FILTRATION AND LONG-TERM RENOPROTECTION

755

As noted, the concept of long-term renoprotection despite initial increases in serum creatinine is not limited to RASI prescription. In the Modification of Diet in Renal Disease Study (MDRD), 7 for example, lower blood pressure It might seem counterintuitive, in the treatment of chronic (without emphasis on RASI use) led to an initial decrease in kidney disease, that interventions glomerular filtration. Despite this that reduce glomerular filtration initial decrease, secondary analycan be beneficial. After all, the sis after 3-year follow-up docuCLINICAL SIGNIFICANCE goal of treatment in chronic kidmented renoprotection in proteinney disease is to preserve glomeruric (⬎1 gram per day) patients in ● Antihypertensive therapies for chronic ular filtration. This seeming parathe lower blood pressure cohort as kidney disease may lead to an initial dox dissolves with an understanding compared with the higher blood increase in serum creatinine, which is of the determinants of glomerular pressure cohort. In addition, post often considered a diagnostic indicator filtration and the difference behoc analysis after an additional tween hemodynamic and parenof treatment failure (“prerenal failure”) 8-year follow-up suggested longchymal changes. and may lead the physician to disconterm renoprotection from lower Glomerular ultrafiltration is regtinue treatment. blood pressure for the entire coulated by four determinants.19 The hort, including patients with ⬍1 ● When increased creatinine results only ultrafiltration coefficient reflects gram of proteinuria.14 from hemodynamic alterations, however, glomerular basement membrane Interestingly, Apperloo et al8 it may in fact indicate successful treatsurface area and permeability and suggested that an initial decrease is the sole determinant pertaining ment that will provide long-term renal in glomerular filtration rate folto parenchymal structure. The other protection and should be continued. lowing blood pressure lowering is three determinants—renal plasma not only acceptable but is prog● We describe this prognostically favorable flow, transcapillary pressure granostically favorable. These austate as “prerenal success.” dient, and oncotic gradient—are thors compared long-term kidney hemodynamic variables. Changes function in patients in whom the in any of the 4 determinants may serum creatinine increased after alter glomerular filtration and the blood pressure control with patients in whom the serum serum creatinine level.19 When RASIs reduce systemic creatinine did not increase. The initial increase in serum and intraglomerular capillary pressure or diuretics recreatinine was associated with long-term renal preservation duce renal plasma flow, glomerular filtration rate may as compared with the patients whose serum creatinine did 9,10 decrease. not increase with blood pressure control, suggesting that When renal function decreases on the basis of scarring or the increase was a biomarker for successful reduction of other parenchymal structural problems and the patient intraglomerular pressure (accounting for long-term renal progresses through the stages of chronic kidney disease, preservation). prognosis declines. However, when the serum creatinine Apperloo et al8 randomized patients to treatment with increases on the basis of hemodynamic alterations, renal either an angiotensin-converting enzyme inhibitor or a betaparenchymal structure and integrity are not immediately blocker. Other antihypertensive medicines, including a diaffected. Similarly, reversing hemodynamic changes with uretic, were added as needed to reach the blood pressure volume expansion or raising blood pressure may decrease goal. The provision of long-term renoprotection via blood the serum creatinine but does not improve parenchymal pressure lowering, despite initial decreases in glomerular structure or function. filtration rate, was demonstrated in both treatment groups, In contrast, renal parenchymal structure and function are 1-7 and these groups included various combinations of antiaffected in the long term by hemodynamic changes. Rehypertensives. Similarly, in the MDRD Study no specific ducing systemic and intraglomerular pressure has been antihypertensive regimen was prescribed.20 A variety of demonstrated to maintain parenchymal function long term, common antihypertensives, including diuretics (50% of even after an immediate decrease in glomerular filtration. In patients in the low blood pressure cohort), were given.20 the Angiotensin-Converting Enzyme Inhibition in ProgresContinuing these medicines even after an initial increase sive Renal Insufficiency Study,2 for example, treatment in serum creatinine in the low blood pressure cohort with the RASI benazepril, with only mild reduction of blood resulted in long-term renoprotection. Although this conpressure, led to higher serum creatinine levels than in the cept is most clearly established for RASI treatment, these placebo arm during the first 6 months of treatment, but studies7,8 suggest that it also applies broadly to blood long-term renal survival, nevertheless, increased. Many pressure control in chronic kidney disease when achieved smaller studies confirm this characteristic of RASI therapy by standard combinations of typically prescribed antihyand provided the basis for a detailed review of this concept pertensive medicines. by Bakris and Weir.10

756

The American Journal of Medicine, Vol 120, No 9, September 2007

It is important to note that there is no pathophysiologic basis to prescribe any a priori limit to an increase in serum creatinine that derives from RASIs, diuresis, and blood pressure lowering, and resulting decreases in renal blood flow or intraglomerular capillary pressure. Based on empirical analysis of the usual results that follow prescription of a single RASI, tolerance to an initial increase in serum creatinine up to 30% (as hemodynamically mediated) has been recommended.9,10,12 These data (and the resulting 30% limit) should not be extrapolated to diuretics. They also do not apply to standard chronic kidney disease therapy, which typically includes the co-administration of diuretics and RASIs, and recently, more aggressive RASI prescription (higher doses and combinations of RASIs) and lower blood pressure goals.11 No defining event occurs as renal blood flow and intraglomerular pressure decrease to any specific point or by any precise percentage in patients with chronic kidney disease and impaired autoregulation.9,19 The degree of decrease in glomerular filtration rate is simply a function of the degree of change in the hemodynamic variables.

A LIMITATION IN THE CURRENT APPROACH TO INCREASES IN SERUM CREATININE The tolerance of physicians for increases in serum creatinine may be limited by our current diagnostic approach. Acute renal failure is defined as a (variable) increase in the serum creatinine. The standard approach to an increase in the serum creatinine is to group potential etiologies into 3 categories of acute renal failure—prerenal, renal, and postrenal. Renal and postrenal etiologies involve immediate or eventual renal parenchymal injury and are always deleterious. In contrast, prerenal etiologies of an increased creatinine involve hemodynamic alterations only, and these hemodynamic alterations, as discussed, often benefit patients with chronic kidney disease. Our current diagnostic algorithm does not distinguish between helpful and harmful prerenal perturbations. Instead, the approach to an increased serum creatinine begins with a presumption that the increased serum creatinine represents a failure and then proceeds by attempting to sort out the reason for the failure. By defining an increased serum creatinine as acute renal failure and conferring upon all prerenal etiologies the same pernicious attributes of renal and postrenal etiologies, we may contribute to the likelihood that a result (increased serum creatinine) of clinically beneficial diuresis, RASI, and blood pressure control will prompt reversal of the beneficial interventions. The absence of a place for beneficial hemodynamic changes in the standard algorithm used to evaluate increases in serum creatinine hinders making this diagnosis. The lack of a formal term for beneficial hemodynamic alterations may be an additional hindrance—it is difficult to diagnose an entity without a discrete name. These problems may be alleviated by modification of the usual approach to acute renal failure (Figure) and coinage of the phrase “prerenal success” (as opposed to prerenal failure) to denote favorable hemodynamic changes (“prerenal

Figure A new diagnostic approach to an increased serum creatinine. This algorithm formally incorporates the diagnosis of prerenal success. When a prerenal etiology to an increased serum creatinine is determined, a broad clinical determination is needed to distinguish prerenal success from prerenal failure.

success” is a variant of the term “acute renal success” used by Thurau and Boylan 21 to describe a different clinical event: a theoretical role of oliguria in reducing salt-wasting in acute tubular necrosis). When an increase in serum creatinine derives from a prerenal etiology, we must not presume that this defines a therapeutic failure that must be reversed. If an overall clinical evaluation of the hemodynamic changes indicates that the patient’s status has improved (as may occur in the treatment of chronic kidney disease), then prerenal success, rather than prerenal failure, should be diagnosed and the favorable therapeutic intervention continued.

ARE THERE DATA THAT ARGUE AGAINST TOLERATING INCREASES IN SERUM CREATININE? Reluctance to tolerating increases in serum creatinine also may derive in part from a developing theme in recent medical literature—the notion that even small increases in serum creatinine are associated with increased morbidity and mortality. The apparent discrepancy between these data and the documentation of improved long-term outcomes in chronic kidney disease needs to be clarified. Small increases in serum creatinine following treatment for decompensated congestive heart failure in hospitalized patients appear to be associated with increases in mortality.22-24 However, these patients have complex, acute, and niche-specific issues that discourage extrapolation from this result to a stable outpatient population with chronic kidney disease. These results may not even be applicable to outpatients with less severe congestive heart failure; for example, increases in serum creatinine in patients given a RASI in the Studies of Left Ventricular Dysfunction (SOLVD) did not preclude continued therapy,25 and these patients “contributed to the primary outcome of the study: improved survival in patients treated with enalapril.”26 In fact, Bart26 suggested that in outpatients with heart failure, the increase in serum creatinine may identify patients

Hirsch

Prerenal Success in Chronic Kidney Disease

who will benefit most from RASIs (as in chronic kidney disease therapy). Moreover, the etiology of increases in serum creatinine in decompensated heart failure is unclear and does not seem related to RASI prescription or negative balance due to diuresis.24 It is possible that neurohumoral mechanisms may be etiologic and that standard therapy of decompensated heart failure simply identifies, via increased serum creatinine, patients with more extensive (but not yet clearly identified) co-morbidities and worse prognosis.27,28 While further research may clarify these issues, cardiology authorities continue to recommend diuretics to relieve pulmonary congestion27 and RASIs for systolic dysfunction,25-28 in the interests of improving overall clinical status and prognosis. In similar fashion, recent data demonstrate that small increases in serum creatinine in various segments of hospitalized patients also are associated with increased mortality.29-32 Several of these studies were limited either to a specific population, coronary bypass patients,29 which is distant from the outpatient chronic kidney disease population, or an etiology, radiocontrast,30 which results in parenchymal disease. Two larger, more general reviews of hospitalized patients with acute renal failure also found that small increases in serum creatinine presaged increased mortality.31,32 Chertow et al32 attempted to distinguish prerenal from other etiologies. However, this was not the primary focus of the study, the etiologic determinations were retrospective, based largely on diagnostic codes without access to clinical or physiologic measurements, and, therefore, of uncertain accuracy. In addition, all prerenal diagnoses were designated as prerenal failure. The types of hemodynamic changes associated with prerenal success could not be isolated and analyzed separately. These studies all make important contributions. However, they do not provide evidence that the poorer prognosis related to various parenchymal diseases or deleterious hemodynamic changes (prerenal failure) in hospitalized patients may be extrapolated to the hemodynamic changes that characterize prerenal success in a different cohort—the stable patient with chronic kidney disease. The strongest evidence against such an extrapolation is the direct analysis of the issue in chronic kidney disease, with multiple studies supporting the concept of prerenal success in these patients.1-8

CONCLUSION: A REVISED APPROACH TO INCREASES IN SERUM CREATININE When the serum creatinine increases, the diagnosis of acute renal failure should be considered but not assumed. Renal parenchymal and postrenal etiologies of acute renal failure should be sought through standard strategies, including ultrasonography (evaluating for hydronephrosis) and urinalysis (with proteinuria, white, red or tubular cells, or cellular casts suggesting parenchymal disease). Prior medicines and procedures should be reviewed to identify exposure to renal toxins. In the absence of evidence of parenchymal and

757 postrenal disease, a prerenal etiology may be considered. This diagnosis would be supported by an increased serum urea nitrogen to creatinine ratio and a reduced fractional excretion of sodium. When the increased serum creatinine appears to have derived from altered hemodyamics, the physician should ask, from a position of equipoise, whether this represents prerenal failure or prerenal success. The determination of success versus failure requires a focus on the resulting hemodynamic and clinical state: Has the patient reached a new hemodyamic state that is well tolerated and associated with improved prognosis? Or, has some particular aspect developed that is injurious to the patient’s well-being? Prerenal success may be diagnosed when hypertensive, proteinuric patients with chronic kidney disease are treated with RAS inhibitors and diuretics with successful resolution of hypertension and proteinuria. For example, if aggressive therapy improves a patient’s systolic blood pressure from 180 mm Hg to 125 mm Hg and reduces his proteinuria from 3 grams per day to 500 mg per day, the therapy should be continued in the interests of improved long-term renal prognosis, even if the serum creatinine has increased. Importantly, the proper diagnosis of prerenal success requires that the increased serum creatinine reach a plateau when the new hemodynamic state stabilizes, generally within 2-4 weeks (although ongoing diuresis, with hemodynamics still in flux, would be an exception).10 Hospitalized patients with a presumptive diagnosis of prerenal success should have the serum creatinine checked daily (outpatients perhaps weekly), until it has stabilized. If the serum creatinine is still increasing after this time period or if it increases again after a plateau phase had been established, prerenal success will be an unlikely diagnosis. However, if the serum creatinine reaches a plateau within 2-4 weeks, and the patient’s hemodynamic condition at this plateau is well tolerated clinically and associated with improved long-term prognosis, then prerenal success may be diagnosed. As noted, in clinical studies demonstrating the long-term benefit of blood pressure control or specific RAS inhibition, RASIs and diuretics were frequently co-administered and were not discontinued based solely on early increases in serum creatinine. The diagnosis of prerenal failure and the subsequent discontinuation or reduction of these medicines requires identifying a specific unfavorable aspect of the hemodynamic change— generally, excessive blood pressure lowering or excessive diuresis. The optimal blood pressure for chronic kidney disease is controversial; criteria for the diagnosis of excessive blood pressure lowering may vary based on individual patient characteristics and, importantly, on whether the specific blood pressure obtained is clinically tolerated. A systolic blood pressure of ⬍125 mm Hg is now typically advised for patients with proteinuria ⬎1 gram per day,11,33,34 and meta-analysis (arguably) suggests that systolic blood pressure as low as 110-119 mm Hg may be renoprotective.13 Systolic blood

758

The American Journal of Medicine, Vol 120, No 9, September 2007

pressure ⬍110 mm Hg may be harmful to long-term kidney function13 and should usually be considered too low. Of note, more conservative blood pressure control is suggested for the elderly, who may not tolerate lower blood pressures. For example, in the MDRD Study the goal in the low blood pressure group was increased from a mean pressure of 92 mm Hg (⬃125/75) to 98 mm Hg (⬃ 130/80) for patients 61 years and older.20 Furthermore, major clinical trials of renoprotection excluded patients over 70 years old;1-7 extrapolation of data endorsing aggressive RASI and blood pressure lowering from younger to older patients may not be warranted.35,36 An individual patient may not tolerate a low blood pressure that is thought to provide optimal renoprotection— symptoms of hypotension mandate reduction of antihypertensive therapy. These symptoms may include dizziness, tachycardia, weakness, oliguria, or other signs of reduced tissue perfusion. In addition, excessive diuresis may manifest with cramps, poor tissue turgor, and significant electrolyte abnormalities. Whenever excessive blood pressure lowering or excessive diuresis is diagnosed, measures should be taken to restore optimal hemodynamic status, including reduction or discontinuation of diuretics or RASIs, and possibly provision of parenteral saline. In summary, hemodynamic changes that increase the serum creatinine may be either helpful or harmful, a distinction that should be made on clinical grounds. Formally naming beneficial hemodynamic changes “prerenal success” and inserting this term in the usual diagnostic algorithm for increases in serum creatinine should facilitate, when appropriate, the diagnosis of beneficial hemodynamic changes and prevent the unnecessary reversal of beneficial therapies.

ACKNOWLEDGMENT The author would like to thank Alan Hirsch for editing the manuscript.

References 1. Lewis E, Hunsicker L, Bain R, Rohde R. The effect of angiotensinconverting-enzyme inhibition on diabetic nephropathy. The Collaborative Study Group. N Engl J Med. 1993;329:1456-1462. 2. Maschio G, Alberti D, Janin G, et al. Effect of the angiotensinconverting enzyme inhibitor benazepril on the progression of chronic renal insufficiency. The Angiotensin-Converting Enyzme Inhibition in Progressive Renal Insufficiency Study Group. N Engl J Med. 1996; 334:939-945. 3. The GISEN Group (Gruppo Italiano di Studi Epidemiologici in Nefrologia). Randomised placebo-controlled trial of effect of ramipril on decline in glomerular filtration rate and risk of terminal renal failure in proteinuric, non-diabetic nephropathy. Lancet. 1997;349:1857-1863. 4. Brenner B, Cooper M, De Zeeuw D, et al. Effects of losartan on renal and cardiovascular outcomes in patients with type 2 diabetes and nephropathy. N Engl J Med. 2001;345:861-869. 5. Lewis E, Hunsicker L, Clarke W, et al. Renoprotective effect of the angiotensin-receptor antagonist irbesartan in patients with nephropathy due to type 2 diabetes. N Engl J Med. 2001;345:851-860. 6. Hou F, Zhang X, Zhang G, et al. Efficacy and safety of benazepril for advanced chronic renal insufficiency. N Engl J Med. 2006;354:131-140.

7. Klahr S, Levey A, Beck G, et al. The effects of dietary protein restriction and blood-pressure control on the progression of chronic renal disease. N Engl J Med. 1994;330:877-884. 8. Apperloo A, De Zeeuw D, De Jong P. A short-term antihypertensive treatment-induced fall in glomerular filtration rate predicts long-term stability of renal function. Kidney Int. 1997;51:793-797. 9. Palmer B. Renal dysfunction complicating the treatment of hypertension. N Engl J Med. 2002;347:1256-1261. 10. Bakris G, Weir M. Angiotensin-converting enzyme inhibitor-associated elevations in serum creatinine. Arch Intern Med. 2000;160: 685-693. 11. Hirsch S. Preventing end-stage renal disease: flexible strategies to overcome obstacles. Curr Opin Nephrol Hypertens. 2006;15:473-480. 12. National Kidney Foundation K/DOQI: clinical practice guideline on hypertension and antihypertensive agents in chronic kidney disease. Am J Kidney Dis. 2004;43(Suppl 1):S120-S141, S183-S222. 13. Jafar T, Stark P, Schmid C, et al for the AIPRD Study Group. Progression of chronic kidney disease: the role of blood pressure control, proteinuria, and angiotensin-converting enzyme inhibition. A patient-level meta-analysis. Ann Intern Med. 2003;139:244-252. 14. Sarnak M, Greene T, Wang X, et al. The effect of a lower target blood pressure on the progression of kidney disease: long-term follow-up of the Modification of Diet in Renal Disease Study. Ann Intern Med. 2005;142:342-351. 15. Vasavada N, Agarwal R. Role of excess volume in the pathophysiology of hypertension in chronic kidney disease. Kidney Int. 2003;64: 1772-1779. 16. De Nicola L, Minutolo R, Bellizzi V, et al. Achievement of target blood pressure levels in chronic kidney disease: a salty question? Am J Kidney Dis. 2004;43:782-795. 17. Bakris G, Williams M, Dworkin L, et al. Preserving renal function in adults with hypertension and diabetes: a consensus approach. National Kidney Foundation Hypertensive and Diabetes Executive Committees Working Group. Am J Kidney Dis. 2000;36:646-661. 18. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: the JNC 7 report. JAMA. 2003;289:2560-2572. 19. Brenner B, Dworkin L, Ichikawa I. Glomerular filtration. In: Brenner B, Rector F, eds. The Kidney, 3rd ed. Philadelphia, PA: W.B. Saunders Company; 1986:127-130. 20. Lazarus J, Bourgoignie J, Buckalew V, et al. Achievement and safety of a low blood pressure goal in chronic kidney disease. The Modification of Diet in Renal Disease Study Group. Hypertension. 1997;29:641-650. 21. Thurau K, Boylan J. Acute renal success. The unexpected logic of oliguria in acute renal failure. Am J Med. 1976;61:308-315. 22. Weinfeld M, Chertow G, Stevenson L. Aggravated renal dysfunction during intensive therapy for advanced chronic heart failure. Am Heart J. 1999;138:285-290. 23. Forman D, Bulter J, Wang Y, et al. Incidence, predictors at admission and impact of worsening renal function among patients hospitalized with heart failure. J Am Coll Cardiol. 2004;43:62-67. 24. Butler J, Forman D, Abraham W, et al. Relationship between heart failure treatment and development of worsening renal function among hospitalized patients. Am Heart J. 2004;147:331-338. 25. Knight E, Glynn R, McIntyre K, et al. Predictors of decreased renal function in patients with heart failure during angiotensin-converting enzyme inhibitor therapy: results from the studies of left ventricular dysfunction (SOLVD). Am Heart J. 1999;138:849-855. 26. Bart B. Concern for azotemia with angiotensin-converting enzyme inhibitors: public health implications and clinical relevance. Am Heart J. 1999;138:801-803. 27. Geisberg C, Butler J. Addressing the challenges of cardiorenal syndrome. Cleve Clin J Med. 2006;73:485-491. 28. Bart B, Goldsmith S. Aggravated renal dysfunction and the acute management of advanced chronic heart failure. Am Heart J. 1999;138: 200-202. 29. Lassnigg A, Schmidlin D, Mouhieddine M, et al. Minimal changes of serum creatinine predict prognosis in patients after cardiotho-

Hirsch

Prerenal Success in Chronic Kidney Disease

racic surgery: a prospective cohort study. J Am Soc Nephrol. 2004;15:1597-1605. 30. Levy E, Viscoli C, Horwitz R. The effect of acute renal failure on mortality. A cohort analysis. JAMA. 1996;275:1489-1494. 31. Uhino S, Bellomo R, Goldsmith D, et al. An assessment of the RIFLE criteria for acute renal failure in hospitalized patients. Crit Care Med. 2006;34:1913-1917. 32. Chertow G, Burdick E, Honour M, et al. Acute kidney injury, mortality, length of stay, and costs in hospitalized patients. J Am Soc Nephrol. 2005;16:3365-3370.

759 33. Ruggenenti P, Schieppate A, Remuzzi G. Progression, remission, regression of chronic kidney diseases. Lancet. 2001;357:1601-1608. 34. Zandi-Nejad K, Brenner B. Strategies to retard the progression of chronic kidney disease. Med Clin North Am. 2005;89:489-509. 35. Hemmelgarn B, Zhang J, Manns B, et al. Progression of kidney dysfunction in the community-dwelling elderly. Kidney Int. 2006;69: 2155-2161. 36. Locatelli F, Pozzoni P. Chronic kidney disease in the elderly: is it really a premise for overwhelming renal failure? Kidney Int. 2006;69: 2118-2120.

The American Journal of Medicine (2007) 120, 760-763

OFFICE MANAGEMENT: ALLERGY AND IMMUNOLOGY Richard D. deShazo, Specialty Editor

Difficult-to-Control Asthma in Adults Forest H. Mealey, DO, Nicholas J. Kenyon, MD, Mark V. Avdalovic, MD, Samuel Louie, MD Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, University of California, Davis, Sacramento. ABSTRACT Difficult-to-control asthma in adults is under-diagnosed and under-treated in the United States, particularly in those 40 years of age or older. Increasing attention has been focused on the subset of adult patients with poorly controlled asthma because they consume up to 85% of all health care dollars spent on asthma, while representing 20% of all asthma patients. In this article, we define difficult-to-control asthma and discuss the problem of misdiagnosis. We present an algorithm for confirming the diagnosis in these patients and highlight the importance of patient self-assessment. © 2007 Elsevier Inc. All rights reserved. KEYWORDS: Misdiagnosis; Severe asthma; Refractory asthma

Asthma affects an estimated 26 million people in the United States and accounts for 13.9 million office visits per year for adults. The annual cost for asthma is nearly $15 billion, with most costs incurred from hospitalizations and emergency department visits for asthma exacerbations. Eighty percent of the direct costs of asthma come from the approximately 20% of patients with difficult-to-control asthma. In this article, we will define and outline the management of difficult-to-control asthma in adults 40 years of age and older and emphasize common problems, such as misdiagnosis.

DEFINITIONS Difficult-to-control asthma or refractory asthma encompasses any or all of the following asthma characteristics: the failure to achieve asthma control with maximal daily doses of inhaled corticosteroids (ICS), the repeated need for emergency health care services, near-fatal attacks, or the need for daily oral corticosteroid therapy (Table 1).1 The European Network for Understanding Mechanisms of Severe Asthma (ENFUMOSA) studied 158 severe asthmatics and found that they were more likely to be female (female/male ratio 4.4:1).2 Since 1999, 75% of more than 500 adults referred to Requests for reprints should be addressed to Nicholas Kenyon, MD, Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, University of California, Davis, 4150 V Street, PSSB Suite 3400, Sacramento, CA 95817. E-mail address: [email protected]

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2006.10.015

our University of California Davis Asthma Network (UCAN) clinics for difficult-to-control asthma are women.

ASTHMA AND THE OFFICE PRACTICE A major goal of an office practice strategy for asthma is to improve the patient’s control of their disease. To do this, every practice should try to implement the Ten Key Clinical Activities for quality asthma care (Table 2).3 These key activities focus on patient self-assessment and education. We find that teaching of inhaler technique and peak flow monitoring is perhaps best performed by nationally certified asthma educators (www.naecb.org), eg, Registered Respiratory Therapists, which we employ in UCAN (Figure 1). Furthermore, every asthma care visit should include a patient-centered assessment of their disease. The Asthma Control Test (ACT) is a 5-question survey that is surprisingly responsive to changes in the patient’s condition (Table 3). A score of ⬍19 on the ACT suggests that patients are uncontrolled with sensitivity of 71% and specificity of 71%.4 In a large European study, the mean per-patient annual cost of asthma management for patients with an ACT ⬍15 was $2005, whereas for patients with an ACT score of ⱖ20, the annual cost was $290.

TESTING IN DIFFICULT-TO-CONTROL ASTHMA Spirometry is recommended in all patients with the diagnosis of asthma. Most severe asthmatics develop a greater

Mealey et al

Adult Severe Asthma

degree of airflow obstruction and less reversibility to bronchodilators than milder asthmatics over time. Among both men and women, and smokers and nonsmokers, adult asthmatics had greater decreases (over 15 years) in forced expired volume in 1 second (FEV1) than those without asthma. Full pulmonary function testing can confirm the diagnosis of asthma, as well as suggest other disease processes, such as chronic obstructive pulmonary disease (COPD), vocal cord dysfunction, or interstitial lung disease. The methacholine challenge test is very useful in establishing the presence or absence of abnormal bronchial hyperresponsiveness; however, it is a relative contraindication to perform the test in patients who have a FEV1 ⬍60% predicted. A negative methacholine challenge test excludes asthma with 95% certainty. Testing exhaled breath samples for nitric oxide (NO) levels is of particular interest in severe asthma. Exhaled NO is increased in asthmatic patients and NO levels correlate with improvement and deterioration in symptoms. Additionally, NO does not appear to be increased in patients with COPD, making the differentiation between these conditions easier. Adoption of this technique will be easier with the advent of handheld NO monitors.

MISDIAGNOSIS Misdiagnosis of difficult-to-control asthma is an important issue, but there are few estimates as to its prevalence in

Table 1 Refractory Asthma*: ATS Workshop Consensus for Clinical Features† Major Characteristics In order to achieve control to a level of mild-moderate persistent asthma: 1. Treatment with continuous or near continuous (ⱖ50% of year) oral corticosteroids 2. Requirement for treatment with high-dose inhaled corticosteroids a. Budesonide ⬎1200 ⬎6 puffs b. Flunisolide ⬎2000 ⬎8 puffs Minor Characteristics 1. Requirement for daily treatment with a controller medication in addition to inhaled corticosteroids, eg, long-acting ␤-agonist, theophylline, or leukotriene antagonist 2. Asthma symptoms requiring short-acting ␤-agonist use on a daily or near daily basis 3. Persistent airway obstruction (FEV1 ⬍80% predicted; diurnal PEF variability ⬎20%) 4. One or more urgent care visits for asthma per year 5. Three or more oral steroid “bursts” per year 6. Prompt deterioration with ⱕ25% reduction in oral or inhaled corticosteroid dose 7. Near-fatal asthma event in the past ATS ⫽ American Thoracic Society. *Requires that other conditions have been excluded, exacerbating factors treated, and patient felt to be generally adherent. †Definition of refractory asthma requires one or both major criteria and 2 minor criteria.

761 primary care office practices. It must be evaluated methodically (Table 4). For example, it is common to misdiagnose asthma as COPD in patients over 40 years of age. Smoking occurs in approximately 30% of asthma patients, and it is common in patients over age 40 to have an overlap syndrome. Besides COPD, other diseases are misdiagnosed as difficult-to-control asthma, including bronchiectasis, bronchiolitis obliterans, aspiration, and vocal cord dysfunction. Furthermore, undiagnosed co-morbid conditions that appear to worsen asthma symptoms include gastroesophageal reflux (GERD), vocal cord dysfunction, and rhinosinusitis. Finally, coronary artery disease (CAD) is common in the aging adult population and can be mistaken for difficult-to-control asthma. The incidence of CAD in asthmatics is not known but likely reflects that in the general adult population. In the ENFUMOSA study, aspirin use was found to be statistically more common in severe asthmatics compared with mild asthmatics.

TREATMENT TRIALS AND ERRORS Apart from daily inhaled corticosteroids, the optimal therapy for difficult-to-control asthma in adults is controversial. No study shows clearly that any drug therapy considerably alters the natural course of asthma in adults. The NIH–

Table 2 The 10 Key Clinical Activities for Quality Asthma Care 1. Establish asthma diagnosis – Rule out other conditions! 2. Classify severity of asthma – Look for a change in the NIH-NAEPP Classification 3. Schedule routine follow-up care See patients q 1-6 months Spirometry q 1-2 years unless unstable Review asthma action plan 4. Assess for referral to specialty care – Know when to refer to asthmatologists 5. Recommend measures to control asthma triggers No smoking Avoid environmental and occupational triggers 6. Treat or prevent co-morbid conditions Rhinosinusitis Vocal cord dysfunction GERD, COPD, flu vaccine 7. Prescribe medications according to severity 8. Monitor use of beta 2 agonists ⬎1 Albuterol MDI/month is cause for alarm 9. Develop a written asthma action plan – Reduces mortality! 10. Provide education on patient self-management Repetition key to learning! Emphasize environmental controls, inhaler techniques, pharmacotherapy Written asthma action plan NIH ⫽ National Institutes of Health; NAEPP ⫽ National Asthma Education and Prevention Program; GERD ⫽ gastroesophogeal reflux disease; COPD ⫽ chronic obstructive pulmonary disease; MDI ⫽ metered dose inhaler.

762

The American Journal of Medicine, Vol 120, No 9, September 2007 Table 4 Checklist of Questions to Ask in Difficult-toControl Asthma ● Is it definitely asthma? Review spirometry. ● Is their a ⬎12% improvement in FEV1 after albuterol? ● Is a methacholine bronchial challenge (PC20) needed to rule out asthma? ● Is the chest X-ray normal? Have other diagnoses been considered? ● Has the correct use of MDIs, DPIs been taught? ● Has a written action plan been prescribed? ● Is bronchoscopy needed to evaluate larynx, vocal cord, upper airway? ● Is a clinical trial of prednisone for 2 weeks with repeat PFTs needed? FEV ⫽ forced expiratory volume; MDI ⫽ metered dose inhaler; DPI ⫽ dry powder inhaler.

severe asthmatics should have their daily emergency treatment plans detailed in an asthma action plan. The asthma action plan is a contract of care holding the patient responsible to a higher level of self-management skills and the clinician to a more aggressive level of care (Figure 2).

CORTICOSTEROIDS To reemphasize, ICS are the mainstay of therapy in difficult-to-control asthma and should be given to all patients first. They alleviate clinical symptoms, improve pulmonary function, and reduce airway inflammation. Sin and colleagues systematically reviewed the long-term effects of all standard controller asthma medications and found, not surprisingly, that inhaled corticosteroids were most effective.5 Small doses of oral prednisone (1-5 mg or equivalent) may be necessary in some patients who have upper airway obstruction or poor inhaler technique and cannot deliver adequate drug doses to the lower airways. Figure 1 Algorithm for treatment and referral of the difficultto-control asthma patient.

NAEPP (National Institutes of Health–National Asthma Education and Prevention Program) guidelines have provided a valuable approach to the initial treatment of asthma patients, but they do not provide specific recommendations when patients fail to achieve asthma control. The updated NAEPP guidelines will address this issue. The most common cause of poorly controlled asthma is nonadherence to treatment. To improve adherence, all adult Table 3 ● ● ● ● ●

LONG-ACTING BETA-2 AGONISTS Long acting beta-2 agonists (LABAs) are indicated for use primarily as corticosteroid-sparing agents and adjuvant ther-

Checklist of Indicators for Poor Asthma Control*

Has your asthma awakened you at night? Have you needed more reliever medications than usual? Have you needed any urgent medical care? Has your peak flow been below your personal best? Are you participating in your usual physical activities? *Available at: www.ginasthma.org.

Figure 2 Triangle of care illustrating partnership between asthma patient, primary care provider, and asthmatologist.

Mealey et al

Adult Severe Asthma

apy in difficult-to-control asthmatics on ICS. LABAs have been the preferred add-on therapy to ICS; however, the Food and Drug Administration has recently scrutinized this recommendation. The Salmeterol Multicenter Asthma Research Trial (SMART) investigators found an increase in the secondary outcome of respiratory-related deaths alone and life-threatening experiences (odds ratio [OR] 4.4) in adult asthmatics on salmeterol, compared with those not on this medication.6 A recent meta-analysis of 19 trials, heavily weighted by SMART, concluded that LABA use is associated with an increased risk of asthma-related exacerbations and deaths.7 LABAs should be used in severe asthmatic adults who have not adequately responded to ICS alone. In turn, removing LABAs from a successful treatment plan can risk under-treatment and cause poor asthma control.

LEUKOTRIENE INHIBITORS Older patients may benefit greatly from a trial of anti-leukotriene drugs, which can attenuate the effects of leukotrienes not blocked by steroids. Anti-leukotriene drugs can be used as alternatives to long-acting beta-2 agonists in adults. The leukotriene inhibitors, montelukast and zafirlukast, and the 5lipooxygenase inhibitor zileuton improve airway function by blocking the inflammatory effects of leukotrienes. Zileuton inhibits neutrophil chemotaxis, which may be applicable in neutrophil-predominant severe asthma.

763 only sign that a patient has a specific polymorphism— particularly the Arg/Arg genotype at position 16 of the beta-receptor—that could predispose them to react adversely to beta-2 agonists. These patients may respond better to anti-cholinergic bronchodilator therapy. Ipratropium bromide is a good alternative rescue bronchodilator for this 15%-30% of asthmatics.

SUMMARY The evaluation and management of difficult-to-control asthma in adults is becoming commonplace for primary care clinicians and asthmatologists alike. A methodical approach to these patients should encompass a reevaluation of the diagnosis, an assessment for concomitant diseases, the establishment of a patient-tailored management strategy, and the institution of a treatment regimen that emphasizes daily inhaled corticosteroids. The referral of patients to asthmatologists allows for patient teaching and disease management planning that may not be feasible in many clinic settings.

ACKNOWLEDGMENT Shannon Mason has our gratitude for her editorial assistance.

References ANTI-IGE ANTIBODY Omalizumab, a recombinant humanized monoclonal antibody against IgE represents a novel therapeutic approach to allergic asthma. By binding the high affinity receptor for the IgE molecule, omalizumab prevents IgE production by Blymphocytes and sensitization of the mast cell. Treatment has been shown to improve quality of life in difficult-tocontrol asthma, when compared with guideline-directed therapy. Patients who experience two exacerbations requiring a hospitalization in a 12-month period or who fail to respond to oral corticosteroids should be considered for a 6-month trial of omalizumab.

SHORT-ACTING BRONCHODILATORS Short-acting beta-2 agonists should be part of an asthma action plan as rescue therapy for asthma patients unless paradoxical worsening of spirometry or symptoms occurs. Regular use of short-acting beta-agonists, however, should alarm the treating physician that the patient is poorly controlled. Poor response to beta-agonist therapy may be the

1. American Thoracic Society. Proceedings of the ATS workshop on refractory asthma: current understanding, recommendations, and unanswered questions. Am J Respir Crit Care Med. 2000;162:2341-2351. 2. European Network for Understanding Mechanisms of Severe Asthma. The ENFUMOSA cross-sectional European multicentre study of the clinical phenotype of chronic severe asthma. Eur Respir J. 2003;22: 470-477. 3. Williams SG, Schmidt DK, Redd SC, Storms W, National Asthma Education and Prevention Program. Key clinical activities for quality asthma care. Recommendations of the National Asthma Education and Prevention Program. MMWR Recomm Rep. 2003;52:1-8. 4. Schatz M, Sorkness CA, Li JT, et al. Asthma Control Test: reliability, validity, and responsiveness in patients not previously followed by asthma specialists. J Allergy Clin Immunol. 2006;117:549556. 5. Sin DD, Man J, Sharpe H, Gan WQ, Man SFP. Pharmacological management to reduce exacerbations in adults with asthma: a systematic review and meta-analysis. JAMA. 2004;292:367-376. 6. Nelson HS, Weiss ST, Bleecker ER, Yancey SW, Dorinsky PM, the SMART Study Group. The Salmeterol Multicenter Asthma Research Trial: a comparison of usual pharmacotherapy for asthma or usual pharmacotherapy plus salmeterol. Chest. 2006;129:15-26. 7. Salpeter SR, Ormiston TM, Salpeter EE. Meta-analysis: effect of longacting beta-agonists on severe asthma exacerbations and asthma-related deaths. Ann Intern Med. 2006;144:904-912.

The American Journal of Medicine (2007) 120, 764-768

OFFICE MANAGEMENT: ALLERGY AND IMMUNOLOGY Richard D. deShazo, Speciality Editor

Evaluation of the Adult with Suspected Immunodeficiency Antoine E. Azar, MD, Zuhair K. Ballas, MD Division of Allergy/Immunology, Department of Internal Medicine, University of Iowa and the Iowa City VA Medical Center, Iowa City. ABSTRACT Several primary immunodeficiencies may have their initial presentation in adulthood. Although recurrent infections are the hallmark of an underlying immunodeficiency, they need not be the presenting manifestation. This review highlights aspects of infections, as well as noninfectious diseases, that should prompt a high index of suspicion for an underlying immune disorder. The office tests that can be obtained for initial screening and their interpretation are detailed. © 2007 Elsevier Inc. All rights reserved. KEYWORDS: Antibody deficiency; Common variable immunodeficiency; Failure to thrive; Hypogammaglobulinemia; Immune deficiency; Opportunistic organisms; Recurrent infections

This article is not intended as a review of the various immunodeficiencies that can be seen in adults. Our objective here is to provide the primary care physician with a practical approach as to when to suspect, and how to screen for, an immunodeficiency in adult patients. A significant part of the evaluation can be done in the office setting. A detailed history can yield abundant clues as to the nature of a potential immune abnormality. Initial screening tests are relatively simple, inexpensive, and widely available. Abnormal results obtained on initial screening may indicate the nature of the immune disorder or the need for consultation with an allergist/immunologist for further evaluation.

A BRIEF OVERVIEW OF THE IMMUNE SYSTEM The main task of the immune system is the differentiation of “self” from “non-self.” “Non-self” is considered harmful and should be eliminated, whereas “self” is tolerated. Microorganisms are the major “non-self” that the immune system attempts to neutralize on a daily basis. A suboptimal immune system is not able to effectively contain microorganisms, resulting in frequent infections. Also, an abnormal immune system might not be tolerant of “self,” resulting in autoimmune diseases. Some immune disorders, therefore,

Requests for reprints should be addressed to Antoine E. Azar, MD, University of Iowa, Internal Medicine, 200 Hawkins Drive, Iowa City, IA 52242. E-mail address: [email protected]

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2006.12.013

may present with increased incidence of infections and autoimmune diseases. Each major class of microorganisms is recognized predominantly by a specific immune mechanism: Gram-negative bacteria are recognized by neutrophils and macrophages. Neutrophils recognize any of several molecules, such as mannose, that are present on the cell wall of the gram-negative bacteria. Encapsulated bacteria have a polysaccharide capsule, and thus their cell wall molecules are hidden. The immune system responds to these organisms by generating antibodies against the capsule. The antigen–antibody complex then activates the complement cascade. Complement fragments (C3b) and the antigen–antibody complex result in optimal opsonization to the macrophages that then phagocytose and eliminate the bacteria. A properly functioning antibody response and complement cascade are needed for optimal defense against these organisms. Intracellular organisms such as viruses and mycobacteria require T lymphocytes and natural killer (NK) cells for their elimination. The immune response against gram-negative bacteria and encapsulated organisms centers on facilitating phagocytosis. This strategy does not work for intracellular organisms because they thrive inside the cell. T lymphocytes and NK cells are responsible for defense against such organisms and function by activating the macrophages (in the case of mycobacteria) or by directly killing the infected cells (in the case of viruses).

Azar and Ballas Table 1

Adult Immunodeficiencies

765

Selected Pathogens Associated with Immunodeficiency Disorders in Adults

Host Defense Affected

Clinical Examples

Common Pathogens

B cells

Common variable immunodeficiency

T cells

AIDS, chronic mucocutaneous candidiasis

Neutrophils

Neutropenia, chronic granulomatous disease

Complement

Terminal complement deficiency (C5-C9) Early complement deficiency (C2, C4, C3) NK cell deficiency

Encapsulated organisms Streptococcus pneumoniae Haemophilus influenzae Moraxella catarrhalis Giardia Campylobacter Fungi Candida Cryptococcus Pneumocystis Viruses Cytomegalovirus Herpes simplex Varicella zoster Mycobacteria Staphylococcus aureus Aspergillus Nocardia Gram-negative bacteria Escherichia coli Serratia marcescens Burkholderia Pseudomonas Neisseria meningitidis, Neisseria gonorrhea Encapsulated organisms (same as B-cell deficiency) Herpes viruses; human papilloma virus

NK cells

AIDS ⫽ acquired immune deficiency syndrome; NK ⫽ natural killer.

When evaluating a patient with frequent infections, the type of organism responsible for these infections should give a clue as to which aspect of the immune system might not be functioning properly (Table 1).1

PRIMARY VERSUS SECONDARY IMMUNODEFICIENCY Primary immunodeficiency disorders (PIDs) are reported to have an estimated prevalence of 1:10,000 in the general population. This is probably an underestimate because the prevalence of selective immunoglobulin (Ig)A deficiency is estimated at 1:300 to 1:700 (in the United States). Approximately one third of patients with selective IgA deficiency have frequent infections, thus making the prevalence of clinically relevant primary immunodeficiency much more common than 1:10,000. In addition to IgA deficiency, the most common PIDs diagnosed in adulthood include common variable immunodeficiency (CVID) and certain complement deficiencies.2 Conditions leading to secondary immune dysfunction are far more common in adults than PIDs. These should always be ruled out first when a patient is being evaluated for suspected immunodeficiency (Table 2). Infections recurring at the same site may suggest an anatomic abnormality, such as a urethral stricture in a patient with recurrent urinary tract infection.

WHEN TO SUSPECT AN IMMUNODEFICIENCY Although recurrent infections are a hallmark of immunodeficiency, this is not always the case. In thinking about immunodeficiency, the key word is “unusual” (Table 3). An unusual organism, an unusual duration, an unusual complication, or an unusual severity of a “usual” infection should

Table 2

Common Causes of Secondary Immunodeficiency

● ● ● ● ●

Malnutrition HIV Malignancy Immunosuppressive drugs Immunomodulatory agents: ● Rituximab (affecting B cells) ● Infliximab, etanercept, adalimumab, anakinra (affecting cellular immunity) ● Drug-induced hypogammaglobulinemia: ● Certain antiepileptics (eg, diphenylhydantoin, carbamazepine, valproate) ● Protein loss (especially if presenting with low IgG but normal IgA and IgM): ● Nephrotic syndrome, protein-losing enteropathy, severe burns ● Metabolic disease: ● Diabetes, severe liver disease, uremia HIV ⫽ human immunodeficiency virus; Ig ⫽ immunoglobulin.

766

The American Journal of Medicine, Vol 120, No 9, September 2007

Table 3 Clues Suggesting the Presence of an Immunodeficiency Aspects of Infections: ● Unusual frequency ● Unusual severity ● Unusual duration ● Unusual complications ● Unusual organisms Noninfectious Clues ● Premature loss of dentition ● Poor wound healing ● Unexplained bronchiectasis ● Chronic diarrhea or malabsorption ● Autoimmunity, especially if more than one (eg, hypothyroidism and alopecia or vitiligo) ● Hematologic disorders (hemolytic anemia, neutropenia, thrombocytopenia) ● “Failure to thrive”

prompt suspicion of an immunodeficiency. For example, an upper respiratory tract infection that progresses to pneumonia or empyema in an otherwise healthy individual may indicate an underlying immune dysfunction. The frequency of infection need not be on a monthly or annual basis. One documented pneumonia every 5 years in an otherwise healthy, nonsmoking, young adult might be considered an increased frequency of infection.3 Certain noninfectious presentations might suggest an underlying immunodeficiency. Premature loss of dentition or recurrent gingivitis may occur in individuals with antibody deficiency or impaired phagocytic function. Poor or delayed wound healing can be indicative of antibody deficiency. Skin or genital warts, especially when extensive and disseminated, suggest a deficient cellular immunity. Immune deficiency disorders also are frequently associated with autoimmune conditions, such as thyroiditis, autoimmune hemolytic anemia, thrombocytopenia, neutropenia, pernicious anemia, inflammatory bowel disease, celiac disease, vitiligo, and connective tissue disorders. A family history of immunodeficiency, frequent infections, or autoimmune disorders also may heighten the suspicion of an immune deficiency.

EVALUATION OF THE PATIENT WITH SUSPECTED IMMUNODEFICIENCY History As in any disease investigation, a thorough history is the most important initial step in evaluation. Detailed information should be obtained about the nature and site of the infections, their frequency, their complications, their documentation (cultures, imaging), and their treatment. Treatment features that might suggest a less than optimal host defense include the need to use intravenous antibiotics, the need for multiple antibiotics, and the need for multiple courses of antibiotics.

History should be obtained regarding premature loss of dentition, poor or delayed wound healing, warts, chronic diarrhea, and bronchiectasis. A history of recurrent aphthous ulcers may be the only clue for the presence of neutropenia. Current and recent medications, especially immunosuppressive and antiepileptic drugs, should be reviewed in detail (Table 2).

Family History Many PIDs are inherited in an autosomal recessive or Xlinked pattern. Consanguinity raises the possibility of an autosomal recessive disorder. CVID and selective IgA deficiency do not have a definitive pattern of inheritance but tend to cluster in families. It is not unusual to have a patient with CVID whose sibling may have IgA deficiency, rheumatoid arthritis, systemic lupus erythematosus, pernicious anemia, or other autoimmune disorders. A small subset of patients with CVID have an autosomal recessive mutation.

Physical Examination Physical examination gives valuable hints to the underlying disease process. Examination of the head and neck may reveal signs of sinusitis, which is common in patients with antibody defects. Tympanic membrane scarring or perforation suggests recurrent otitis media. Oral examination should include evaluation for aphthous ulcers and oral thrush. Rales or rhonchi may indicate bronchiectasis or chronic bronchitis. Hepatosplenomegaly is seen in a significant subset of patients with CVID. Lymphadenopathy also is common in CVID (reactive lymphadenopathy is the most frequent histologic picture, although lymphoma is also markedly increased). Skin may reveal verrucae, furuncles or abscesses, severe scars (suggestive of possible repeated

Table 4 Initial Screening Tests for Secondary Causes of Immunodeficiency ● CBC with differential and smear: Leukocytosis (infection, leukemia) Neutropenia (medications, infections, malignancy) Lymphopenia (HIV, malignancy) Eosinophilia (allergic disorders, lymphoma) Anemia (chronic disease, malabsorption, hemolysis) Macrocytosis (alcoholism, pernicious anemia) Thrombocytopenia (bone marrow infiltration, autoimmunity) ● Creatinine, electrolytes, liver function tests, blood glucose ● Urinalysis (proteinuria) ● Total serum protein, albumin, globulin ● HIV (ELISA or PCR) ● Imaging to document sinusitis or bronchiectasis, if indicated ● Appropriate cultures if indicated CBC ⫽ complete blood count; HIV ⫽ human immunodeficiency virus; ELISA ⫽ enzyme-linked immunosorbent assay; PCR ⫽ polymerase chain reaction.

Azar and Ballas Table 5

Adult Immunodeficiencies

767

Immunologic Screening Tests

Office-based Immunologic Screening Tests

● B cells Quantitative immunoglobulins (IgG, IgA, IgM, IgE)* Flow cytometry: enumeration of B cells Antibody response to pneumococcus, tetanus, and diphtheria vaccines ● T cells Flow cytometry: enumeration of T cells (CD3), T-helper cells (CD4), and cytotoxic T cells (CD8) ● Complement Total hemolytic complement (CH50) Selected complement levels as suggested by history ● Neutrophils/phagocytes CBC with differential Respiratory burst generation by flow cytometry (available in many reference laboratories) ● NK cells Flow cytometry: enumeration of NK cells (CD16/CD56) CBC ⫽ complete blood count; Ig ⫽ immunoglobulin; NK ⫽ natural killer. *IgG subclasses are not recommended as a screening test.

pyogenic infections), or vitiligo. Joint deformities suggest previous joint infections, or autoimmune inflammatory conditions, seen in antibody and complement deficiencies.4

Initial Diagnostic Evaluation After the history and physical examination, routine screening laboratory tests provide helpful information (Table 4). It is important to note that patients with a defect in antibody production may have false-negative results on tests that involve the measurement of antibodies. For example, to rule out hepatitis C or human immunodeficiency virus (HIV) infection in a patient with antibody deficiency, a viral load should be obtained rather than enzyme-linked immunosorbent assay. Microbiological information is critical for indicating the abnormal aspect of the immune system (Table 1).

Table 6

Initial immunologic screening tests are readily available and can be done in an office setting (Table 5).

Antibody Deficiency Quantitative evaluation of the humoral immune system starts with measurement of serum Igs (IgG, IgA, IgM, and IgE) (Table 6). An elevated IgE level might indicate further evaluation for allergic disorders. If the Ig levels are normal in the face of a strong clinical suspicion of an antibody deficiency, an assessment of functional activity of B cells should be undertaken. Some patients may have normal Ig levels (probably produced by memory B cells that developed before the acquisition of the immune abnormality) but are unable to mount a specific antibody on challenge. The definitive method for functional assessment of B cells consists of measuring specific antibodies after immunization. Both polysaccharide and protein vaccines are used. Antibodies should be measured before vaccine administration and repeated 4 weeks later. It is important to check the antibody levels before vaccination, because the interpretation of the results depends on the fold increase in antibody titers. Definitive criteria for an adequate antibody response have not been established, but a generally accepted rule is a 4-fold increase in antibody titers. Interpretation becomes more difficult in patients who have protective titers before immunization and who subsequently may not adequately respond to the vaccine. Common protein vaccines used are tetanus and diphtheria; a common polysaccharide vaccine used is the 23valent pneumococcal vaccine, which provides the potential for measuring the immune response to different pneumococcal serotypes. The consensus is that normal individuals mount a 4-fold antibody titer increase in response to 70% of pneumococcal serotypes.2 The utility of measuring IgG subclasses is questionable. Although IgG subclasses can be selectively low,

Interpretation of Serum Immunoglobulin Levels*

Immunoglobulin Level IgG

IgA

IgM

Differential Diagnosis

Nl

2

Nl

Nl or 1

2

2

2

Nl

Nl

2 2

2 or Nl 2

2 or Nl Nl or 1

Selective IgA deficiency (if IgA is nondetectable) Anti-epileptics Monoclonal IgG production Common variable immunodeficiency patient on replacement therapy Protein loss (nephrotic syndrome, protein-losing enteropathy) Antibody deficiency Common variable immunodeficiency (low IgA and/or low IgM) Monoclonal gammopathy (Waldenstrom, monoclonal gammopathy of unknown significance, multiple myeloma) Hyper-IgM syndrome

Nl ⫽ normal; 2 ⫽ decreased; 1 ⫽ increased; Ig ⫽ immunoglobulin. *Any elevated immunoglobulin level should be followed by immunofixation electrophoresis to look for a monoclonal band.

768

The American Journal of Medicine, Vol 120, No 9, September 2007

one should not make any therapeutic decision on the basis of the level alone. Individuals with chromosomal deletions of IgG2 have been identified and had no clinical disease, probably because they were able to compensate by shifting their IgG response to the other subclasses. Current recommendations focus on evaluating specific antibody response with post-immunization titers rather than measuring IgG subclasses.

Cellular Immunity T-cell abnormalities result in infections with viral, fungal, or opportunistic organisms. Initial screening should always include HIV testing and enumeration of lymphocyte subsets. The cells tested are CD3 (T cells), CD4 (helper T cells), CD8 (cytotoxic T cells), CD19 or CD20 (B cells), and CD16 or CD56 (NK cells). Although an abnormal CD4:CD8 ratio may be important in HIV, its significance in the absence of HIV is not well established. A low CD4 count with negative HIV testing might indicate idiopathic CD4 lymphopenia, and the patient should be referred to an allergist/immunologist for further assessment.5 Delayed-type hypersensitivity is the major in vivo screening test for T-cell dysfunction. This is performed by using various antigens to which the patient has been exposed. However, administration and interpretation of such tests are fraught with difficulty, especially if an office does not routinely perform such tests. A major concern is the proper choice of antigens for the skin tests because there is no one single antigen to which 100% of the population responds. Therefore, several antigens have to be used before one can be confident that a negative test result indicates anergy. Because in vitro lymphocyte testing may be required in place of delayed skin testing, consultation should be considered when specific cell-mediated deficits are suspected.

Complement Patients with early complement component (C2, C4) deficiency present with collagen vascular disease and infections similar to patients with B-cell deficiency. Patients with C3 deficiency may be indistinguishable from those with antibody deficiency. Patients with terminal complement deficiency (C5, C6, C7, C8, C9) present with recurrent Neisserial infections. The most useful test for screening for complement deficiency is total hemolytic complement (CH50), which reflects the activity of all components of the classic complement pathway (C1 to C9) and the terminal components of the alternative pathway. If CH50 is low, the patient should be referred to an allergist/immunologist for further evaluation. Screening for abnormalities of the alternative complement pathway

can be done using rabbit eryhrocytes. The most common cause of a low CH50 is inappropriate handling of the serum sample.

Phagocytes Patients with phagocytic cell defects develop recurrent bacterial and fungal infections involving the skin, periodontal tissue, lung, liver, and bone. Neutropenia is the most commonly encountered disorder of phagocytes and is detected by doing complete blood cell count with differential. Most neutropenias are secondary to medications. An infection pattern suggestive of neutrophil deficit in the presence of normal neutrophil count should be followed by testing for superoxide generation (available in most reference laboratories using flow cytometry).

Natural Killer Cells NK cell deficiency, a rare entity, is associated with recurrent herpes viral infections, as well as severe human papilloma virus-associated disease, such as vaginal and cervical cancers or recurrent and disseminated warts. NK cell numbers can be assessed by flow cytometry. Functional assessment is available in reference libraries.

SUMMARY A heightened index of suspicion of immunodeficiency can lead to an early diagnosis and a favorable outcome. Recurrent infections, unusual aspects of usual infections, unexplained bronchiectasis, chronic diarrhea, autoimmune diseases, and “failure to thrive” should prompt screening for a potential immune abnormality. A considerable wealth of information can be obtained from screening tests done in a primary care setting. If initial assessment does not reveal any abnormality, and the clinician still suspects an immunodeficiency, referral to an allergist/immunologist for further evaluation is indicated.

References 1. Holland SM, Gallin JI. Evaluation of the patient with suspected immunodeficiency. In: Mandell GL, Bennett JE, Dolin R, eds. Principles and Practice of Infectious Diseases. 6th ed. New York, NY: Elsevier; 2005:149-160. 2. Bonilla FA, Bernstein IL, Khan DA, et al. Practice parameter for the diagnosis and management of primary immunodeficiency.. Ann Allergy Asthma Immunol2005;94(5 Suppl 1):S1-S63. 3. Kavanaugh A. Evaluation of patients with suspected immunodeficiency. Am Fam Physician. 1994;49:1167-1172. 4. Ballow M, O’Neil KM. Approach to the patient with recurrent infections. In: Adkinson NF, Yunginger JW, Busse WW, Bochner BS, Holgate ST, Simons FE, eds. Middleton’s Allergy Principles and Practice.6th ed. Philadelphia, PA: Mosby; 2003:1043-1072. 5. Fleisher TA, Oliveira JB. Functional and molecular evaluation of lymphocytes. J Allergy Clin Immunol. 2004;114:227-234.

The American Journal of Medicine (2007) 120, 769-771

DIAGNOSTIC DILEMMA: ENDOCRINOLOGY Charles M. Wiener, MD, Section Editor

A Vast Differential Dana L. Madison, MD, PhD Division of Endocrinology, Oregon Health and Sciences University, Portland, Ore.

PRESENTATION For one patient, multiple discrete abnormalities contributed to a singular result. Over an 11-month period, a 39-year-old white woman experienced increasing fatigue, hot flashes, palpitations, nausea, vomiting, and abdominal pain. She gave birth 11 months before her presentation and had ceased lactation 2 months earlier. During the previous 2 months, she also lost roughly 30 pounds (14 kg). The family history was significant for type 2 diabetes mellitus and autoimmune hypothyroidism. She did not smoke and denied use of alcohol or illicit drugs. Her only medication was prenatal vitamins.

(PTH), 6 pg/mL (normal, 15-75 pg/mL); normal 25-hydroxy-vitamin D, 39 ng/mL (normal, 20-57 ng/mL); inappropriately high-normal 1,25 dihydroxy-vitamin D, 71 pg/mL (normal, 15-75 pg/mL); and elevated PTH-related peptide, 11.6 pmol/L (normal, ⬍ 0.3 pmol/L) (Figure 1). A computerized tomography scan showed symmetric, lobulated, hilar adenopathy with prominent paratracheal nodes (largest, 4.0- x 1.9 cm), diffuse interstitial lung disease, peribronchovascular nodules, and subpleural nodularity. Pulmonary function tests indicated that lung volume, lung flow, and diffusion capacity were normal. A transbronchial lung biopsy confirmed noncaseating granulomas, which were negative for acid-fast bacilli and malignant tissue.

ASSESSMENT A physical examination disclosed a regular pulse of 102 beats per minute; blood pressure, 148/85 mm Hg; negative ocular findings; a palpable goiter without discreet nodules; scattered lymph nodes in the supraclavicular and auricular chains; clear lungs; and a fine upper-extremity resting tremor with mild breakaway weakness. Laboratory studies revealed significant hypercalcemia, hyperthyroxinemia, and mild renal insufficiency. Pertinent findings were as follows: total calcium, 15.9 mg/dL (normal, 8.5-10.5 mg/dL); ionized calcium, 2.1 mmol/L (normal, 1.14-1.32 mmol/L); phosphorous, 5.3 mg/dL (normal, 2.4-4.7 mg/dL); thyroid-stimulating hormone,⬍0.01 mIU/L (normal, 0.28-5 mIU/L); free thyroxine, 5.0 ng/dL (normal, 0.7-1.8 ng/dL); total T3, 483 ng/dL (normal, 68-200 ng/dL); creatinine, 1.3 mg/dL (normal, 0.6-1.1 mg/dL); hematocrit, 29% (normal, 36%-46%); and alkaline phosphatase, 121 U/L (normal, 42-98 U/L). A chest x-ray demonstrated expanded lung volumes and no nodules or hilar adenopathy. The patient was admitted for management of significant hypercalcemia and hyperthyroxinemia. Further studies demonstrated suppressed intact parathyroid hormone Requests for reprints should be addressed to Dana L. Madison, MD, PhD, Division of Endocrinology, L-607, Oregon Health and Sciences University, 3181 SW Sam Jackson Park Rd, Portland, OR 97239. E-mail address: [email protected]

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2007.06.002

DIAGNOSIS The findings from computed tomography and transbronchial lung biopsy suggested Stage 3 sarcoidosis (Figure 2). This case illustrates multiple etiologies for PTH-independent hypercalcemia in a single patient. Her severe calcium levels (15.9 mg/dL) occurred when the effects of the primary lesion, sarcoidosis, were compounded by consequences of hyperthyroxinemia and mild renal insufficiency; the last was triggered by hypercalcemia and hyperthyroidism. While it is well-known that hypercalcemia can result when granulomatous 1␣-hydroxylase activity fosters synthesis of excessive 1,25 dihydroxy-vitamin D, the role of PTH-related peptide, which is secreted by granulomatous tissue, is just beginning to be understood as a pathophysiologic mediator for hypercalcemia in these patients. An acute presentation with symptomatic hypercalcemia is rare in sarcoidosis. Significant hypercalcemia is present in only 5-10% of sarcoid patients; hypercalciuria, in 40-62%.1,2,3 Non-PTH-mediated hypercalcemia can result from numerous disorders, including carcinomas (osteolytic lesions and tumors that manufacture PTH-related peptide), medications (eg, thiazides, tamoxifen, calcitriol), immobilization, renal insufficiency, endocrinopathies (eg, hyperthyroidism, adrenal insufficiency), and both infectious (eg, HIV, tuberculosis) and noninfectious (eg, sarcoidosis, Wegener’s) granuloma-forming diseases.

770

The American Journal of Medicine, Vol 120, No 9, September 2007

Figure 1 Levels of total serum calcium, 25-hydroxy-vitamin D, 1,25 dihydroxy-vitamin D, and parathyroid hormone are compared at different points: presentation, during hospitalization, and 10 months after discharge. The corticosteroid treatment for 1 month is shown as well. Total calcium, squares on solid, black, line; 25hydroxy-vitamin D, diamonds on dashed line; 1,25 dihydroxyvitamin D, circles on broken gray line; and parathyroid hormone, diamonds on solid line.

min D.11,12 These data suggest that PTH-independent hypercalcemia in sarcoidosis occurs by dysregulation of macrophage 1␣ -hydroxylase within granulomatous tissue.3,13 Recently, detectable levels of PTH-related peptide were found within about 85% of pulmonary macrophage biopsy specimens, providing another potential stimulus for hypercalcemia.14 A unifying mechanism for hypercalcemia due to elevations in both PTH-related peptide and 1,25 dihydroxyvitamin D is suggested by the macrophage response to specific cytokines. In active sarcoidosis, alveolar cells secrete interleukin-2, recruiting macrophages for activation and differentiation and also increasing local interferon-␥ production. In turn, interferon-␥ induces 1␣ -hydroxylase activity. The result: elevated 1,25 dihydroxy-vitamin D with increases in bone turnover and intestinal calcium absorption. Additionally, these cytokines may also increase PTHrelated peptide secretion.3

MANAGEMENT Moderate to severe hyperthyroxinemia can cause mild hypercalcemia in about 50% of untreated individuals. However, typical serum calcium values are 10.5-11.5 mg/dL, and these suppress PTH and 1,25 dihydroxy-vitamin D. Although the development of sarcoid tissue has been reported in Graves’ glands, this patient’s thyroid gland had a homogeneous appearance on computed tomography, making granulomatous invasion of the gland unlikely.4 Given the variety of potential diagnoses, PTH-independent hypercalcemia requires careful assessment of the calcitropic hormone axis, including evaluation of PTH, 25hydroxy-vitamin D, 1,25 dihydroxy-vitamin D, PTH-related peptide, phosphorus, magnesium, creatinine, and 24-hour urine calcium. In this patient, hypercalcemia combined with an inappropriately elevated 1,25 dihydroxy-vitamin D level and suppressed PTH generated a search for granulomatous and lymphoproliferative disorders, with concern for a malignant process. Key findings illuminated the connection between hypercalcemia and sarcoidosis. Researchers noted that calcium balance in sarcoid patients varied without an increase in calcium intake and that corticosteroids improved hypercalcemia and hyperphosphatemia.5,6 Furthermore, when sarcoid patients with hypercalcemia and low PTH levels took small doses of vitamin D2, their 1,25 dihydroxy-vitamin D levels increased, indicating that a non-PTH mechanism was responsible.7,8 Later, sarcoid macrophages were shown to metabolize 25-hydroxy-vitamin D into 1,25 dihydroxy-vitamin D, confirming that granulomatous tissue was a source of endogenous production.9 The substrate specificity and enzyme kinetics for 1␣ – hydroxylase is similar in macrophages and the kidneys, yet the disease process suggested different regulation.10 Macrophage 1␣ hydroxylase was induced by cytokines, including interferon-␥, but not by PTH or calcitonin. In contrast to renal 1␣ -hydroxylase, it was suppressed by glucocorticoids but not by 1,25 dihydroxy-vita-

Therapy in sarcoidosis with increased 1,25 dihydroxy-vitamin D focuses on acute management of the hypercalcemia and chronic control of the underlying granulomatous disease. Our patient’s hypercalcemia and renal insufficiency was addressed with aggressive hydration. Corticosteroids were administered to reduce the 1,25 dihydroxy-vitamin D level; methimazole and a beta-adrenergic blocking agent were given to control the hyperthyroxinemia symptoms. Her creatinine normalized (0.8 mg/dL) after 36 hours, while the calcium normalized (9.8 mg/dL) after 5 days. Corticosteroids decrease macrophage activation, thereby attenuating synthesis of both 1␣-hydroxylase and PTH-related peptide. Laboratory studies after 1 month’s treatment with a corticosteroid bolus-and-taper, methimazole, and minimized sun exposure revealed the following: serum calcium, 8.9 mg/dL; ionized calcium, 1.18 mmol/L; PTH, 26 pg/mL; 25-hydroxy-

Figure 2 The largest (4.0- x 1.9 cm) hilar lymph node is highlighted with an arrow on this computerized tomography scan.

Madison

A Vast Differential

vitamin D, 45 ng/mL; 1,25 dihydroxy-vitamin D, 47 pg/mL; PTH-related peptide, ⬍ 0.3 pmol/L, thyroid stimulating hormone, 0.19 mIU/L; and free thyroxine, 0.6 ng/dL. These results demonstrated the corticosteroid effect in decreasing both the inappropriately elevated 1,25 dihydroxy-vitamin D and the PTH-related peptide production from granulomatous tissue. At the same time, serum calcium was normalized and accompanied by an appropriate rise in PTH. For long-term management, the patient undergoes periodic calcium measurements and yearly pulmonary function tests with a chest x-ray. She follows a normal calcium diet and remains on methimazole for Graves’ disease. After 18 months and no further corticosteroid treatment, she remains normocalcemic with appropriate levels of PTH and 1,25 dihydroxyvitamin D. Her thyroid function tests remain normal. Complications resulting from both the disease process and corticosteroid use in sarcoid patients with chronic hypercalciuria can produce osteoporosis, nephrolithiasis, and renal insufficiency. Management of the long-term skeletal effects requires careful monitoring for hypercalcemia and hypercalciuria, with treatment of osteoporosis or bone loss in the at-risk patient. In sarcoid patients without hypercalcemia, the risks of bone loss are greatest when significant hypercalciuria is present. Therefore regular assessment of the calcitropic axis and bone metabolism is advisable.

References 1. Lebacq E, Desmet V, Verhaegen H. Renal involvement in sarcoidosis, Postgrad Med J. 1970; 46:526-529. 2. Studdy PR, Bird R, Neville E, James DG. Biochemical findings in sarcoidosis. J Clin Pathol. 1980;33:528-533.

771 3. Porter N, Beynon HL, Randeva HS. Endocrine and reproductive manifestations of sarcoidosis. QJM. 2003;96:553-561. 4. Zimmermann-Belsing T, Christensen L, Hansen HS, et al. A case of sarcoid granuloma, papillary carcinoma, and Graves’ disease in the thyroid gland. Thyroid. 2000;10:275-278. 5. Harrell GT, Fisher S. Blood chemical changes in Boeck’s sarcoid with particular reference to protein, calcium and phosphatase values. J Clin Invest. 1939;18:687-693. 6. Albright F, Carroll EL, Dempsey EF, Henneman PH. The cause of hypercalcuria in sarcoid and its treatment with cortisone and sodium phytate. J Clin Invest. 1956;11:1229-1242. 7. Bell NJ, Gill JR, Jr, Bartter FC. On the abnormal calcium absorption in sarcoidosis. Evidence for increased sensitivity to vitamin D. Am J Med. 1964;36:500-513. 8. Bell NH, Stern PH, Pantzer E, et al. Evidence that increased circulating 1␣, 25-dihydroxyvitamin D is the probable cause for abnormal calcium metabolism in sarcoidosis, J Clin Invest. 1979; 64:218-225. 9. Adams JS, Singer MA, Gacad MA, et al. Isolation and structural identification of 1,25-dihydroxyvitamin D3 produced by cultured alveolar macrophages in sarcoidosis. J Clin Endocrinol Metab. 1985; 60:960-966. 10. Reichel H, Koeffler HP, Bishop JE, Norman AW. 25-hydroxyvitamin D3 metabolism by lipopolysaccharide-stimulated normal human macrophages. J Clin Endocrinol Metab. 1987;64:1-9. 11. Pinkston P, Saltini C, Muller-Quernheim J, Crystal RG. Corticosteroid therapy suppresses spontaneous interleukin 2 release and spontaneous proliferation of lung T lymphocytes of patients with active pulmonary sarcoidosis. J Immunol. 1987;139:755-760. 12. Monkawa T, Yoshida T, Hayashi M, Saruta T. Identification of 25hydroxyvitamin D3 1␣-hydroxylase gene expression in macrophages. Kidney Int. 2000;58:559-568. 13. Conron M, Young C, Beynon HL. Calcium metabolism in sarcoidosis and its clinical implications. Rheumatology (Oxford). 2003;39:707713. 14. Zeimer HJ, Greenaway TM, Slavin J, et al. Parathyroid-hormonerelated protein in sarcoidosis. Am J Pathol. 1998;152:17-21.

The American Journal of Medicine (2007) 120, 772-774

DIAGNOSTIC DILEMMA: GASTROENTEROLOGY Charles M. Wiener, MD, Section Editor

A Rolling Stone Marco A. Gonzalez, MD,a Joseph J. Sorrento, MD,b Ernest Tsao, MDc Department of aSurgery and cDepartment of Medicine, Division of Gastroenterology-Hepatology, Stony Brook University Medical Center, Stony Brook, NY and bDepartment of Surgery, Northport Veterans Affairs Medical Center, Northport, NY.

PRESENTATION In this case, seemingly ordinary symptoms were traced to an unusual source. A 73-year-old male presented to the emergency department with a 5-day history of poorly localized, intermittent abdominal pain that was unrelated to meals. Associated symptoms included nausea, diarrhea, and progressive abdominal distention, which prompted him to seek medical care. He denied any history of fever, chills, vomiting, or unintentional weight loss. The patient’s past medical history was significant for hypertension, coronary artery disease, type-2 diabetes mellitus, and dyslipidemia. His only surgical history was a left inguinal hernia repair. Four weeks earlier, a computed tomography scan of the abdomen and pelvis, performed for an unrelated condition, disclosed asymptomatic cholelithiasis.

ASSESSMENT On physical examination, distention of the patient’s abdomen, hyperactive bowel sounds, and generalized tenderness to palpation were noted. His cardiopulmonary examination was unremarkable, vital signs were normal, and results of laboratory testing, including a white blood cell count, liver function tests, and amylase and lipase levels, were all within normal limits. An abdominal x-ray revealed dilated loops of bowel with air fluid levels plus air in the biliary tree. A computed tomography scan of the abdomen and pelvis demonstrated air within the gallbladder and a calcified intraluminal mass in the descending and sigmoid colon (Figure 1). Of significance was the absence of a large gallstone in the gallbladder, which had been visible on a computed tomography scan that was obtained 4 weeks earlier (Figure 2). Although colonic obstruction from a gallstone is very rare, accounting for only 2% to 8% of all cases of gallstone ileus, the Requests for reprints should be addressed to Ernest Tsao, MD, Stony Brook University Medical Center, Division of Gastroenterology-Hepatology, Health Science Center, T-17, Room 060, Stony Brook, NY 11794. E-mail address: [email protected]

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2007.06.018

diagnosis of colonic gallstone ileus was considered, based on the computed tomography findings.1

DIAGNOSIS Colonoscopy, carried out with informed consent, confirmed the presence of an obstructed gallstone in the sigmoid colon (Figure 3). Gallstone ileus is an unusual source of gastrointestinal obstruction, and it is associated with an 8% rate of morbidity and mortality, the result of delayed diagnosis and patient comorbidities.2 Furthermore, an increased incidence is seen in patients older than 65 years. Many of these patients will have concomitant cardiovascular disease, diabetes, or morbid obesity at the time of diagnosis.1 Gallstone ileus is caused by fistula formation and subsequent stone migration from the gallbladder or less frequently, the common bile duct. Fistula formation occurs when the gallbladder ruptures into adjacent bowel or when a large stone produces direct-pressure necrosis of the gallbladder wall. In the presence of a cholecystoduodenal or cholecystogastric fistula, the most common site of intestinal obstruction is at the terminal ileum. Other sites of obstruction, in order of decreasing frequency, include the proximal ileum, jejunum, and duodenum.2,3 Colonic obstruction can be brought about by a cholecystocolic or cholecystoduodenocolic fistula.3 Blockage usually occurs in the sigmoid colon. These patients typically have concomitant diverticular disease or carcinoma at the point of impaction.4 Among those who have a normal colon without underlying disease—like our patient— obstruction can still occur when the gallstone is greater than 2.5 cm. Smaller stones have been reported to cause obstruction in the presence of fecal impaction.3,4 Patients generally present with intermittent episodes of abdominal pain, vomiting, and diarrhea, reflecting temporary blockage by the stone and ensuing dislodgment, as it traverses distally in the bowel, a phenomenon known as tumbling obstruction.5 Plain abdominal films can be helpful in the diagnosis of gallstone ileus. Classic findings include pneumobilia, evidence of intestinal obstruction, and an ab-

Gonzalez et al

A Rolling Stone

773

Figure 3 Figure 1 Contrast enhanced abdominal computed tomography was performed upon admission. The arrow shows a calcified mass (arrow) in the descending and sigmoid colon, which caused proximal colonic dilatation.

errantly-located gallstone, often seen with the aid of an oral contrast.3,5 Computed tomographic imaging of the abdomen may show pneumobilia and the site of obstruction within the gastrointestinal tract. Laboratory findings are usually nonspecific and might include slight leukocytosis, electrolyte imbalance, and evidence of dehydration.5

MANAGEMENT Multiple attempts to extract the stone endoscopically were unsuccessful, and the patient had to be taken to the operat-

Figure 2 Four weeks before admission, contrast enhanced abdominal computed tomography revealed a large gallstone (arrow) within the gallbladder.

A large gallstone was found during colonoscopy.

ing room for exploratory laparotomy. During the procedure, a serosal tear was observed in the sigmoid colon where the gallstone was lodged. Since the stone could not be manually milked into the rectum, an enterolithotomy was performed with extraction of a large gallstone, measuring 5.8- ⫻ 4.2- ⫻ 3.3 cm (Figures 4 and 5). The gallstone appeared to have passed through a cholecystocolic fistula, which was found between the gallbladder and the ascending colon with surrounding fibrosis and inflammation. After the fistula was repaired, a partial cholecystectomy was performed, along with a right hemicolectomy and a temporary diverting ileostomy. The patient’s postoperative recovery was uneventful, and reversal of the ileostomy was planned. Because spontaneous passage of the stone is rare, emergent surgical intervention remains the treatment of choice for stone extraction and relief of obstruction.3,6 This usually entails enterolithotomy, with or without cholecystectomy, and takedown of the cholecystoenteric fistula.7 Bowel resection is sometimes necessary, particularly in the presence

Figure 4 Emergent laparotomy with enterolithotomy was the procedure of choice.

774

The American Journal of Medicine, Vol 120, No 9, September 2007

References

Figure 5

The gallstone was removed.

of a perforated bowel. There have been reports of successful endoscopic removal of an obstructing colonic gallstone by colonoscopy.8,9 Shock-wave lithotripsy also has been used successfully in treating patients who were not candidates for surgery.10

1. O’Donoghue GT, Winter D, Deasy J. Image of the month. Cholecystocolic fistula and large-bowel obstruction due to gallstone ileus. Arch Surg. 2003;138:1391-1392. 2. Ishikura H, Sakata A, Kimura S, et al. Gallstone ileus of the colon. Surgery. 2005;138:540-542. 3. Clavien PA, Richon J, Burgan S, Rohner A. Gallstone ileus. Br J Surg. 1990;77:737-742. 4. Anseline P. Colonic gall-stone ileus. Postgrad Med J. 1981;57:62-65. 5. Reisner RM, Cohen JR. Gallstone ileus: a review of 1001 reported cases. Am Surg. 1994;60:441-446. 6. Anagnostopoulos GK, Sakorafas G, Kolettis T, et al. A case of gallstone ileus with an unusual impaction site and spontaneous evacuation. J Postgrad Med. 2004;50:55-56. 7. Pavlidis TE, Atmatzidis KS, Papaziogas BT, Papaziogas TB. Management of gallstone ileus. J Hepatobiliary Pancreat Surg. 2003; 10:299-302. 8. Roberts SR, Chang C, Chapman T, et al. Colonoscopic removal of a gallstone obstructing the sigmoid colon. J Tenn Med Assoc. 1990;83: 18-19. 9. Zaretzky B, Kodsi BE, Iswara K. Colonoscopic diagnosis and relief of large bowel obstruction caused by impacted gallstone. Gastrointest Endosc. 1977;23:210-211. 10. Holl J, Sackmann M, Hoffmann R, et al. Shock-wave therapy of gastric outlet syndrome caused by a gallstone. Gastroenterology. 1989; 97:472-474.

The American Journal of Medicine (2007) 120, 775-777

IMAGES IN DERMATOLOGY Parwathi “Uma” Paniker, MD, Section Editor

A Pickled Plum of a Problem Axel Trautmann, Eva-B. Bröcker, Cornelia S. Seitz Department of Dermatology, Venerology and Allergology, University of Würzburg, Würzburg, Germany.

PRESENTATION

DIAGNOSIS

An attempt to lose weight can have unexpected consequences, but a rash is surely one of the least expected. In this case, a 38-year-old female was referred by her family doctor because of spreading skin eruptions that had started approximately 6 weeks prior with single lesions on the arms. She had been diabetic (type II) for 5 years and was currently being treated with metformin, repaglinide, and insulin. For the last 8 weeks she had been attempting to lose weight on a diet that included umeboshi, a very salty pickled Japanese plum and had therefore increased her intake of carbohydrate-rich drinks. She also complained of headache and dizziness but experienced no gastrointestinal symptoms.

Xanthomas are frequently associated with disorders of lipid metabolism. Since these disorders can lead to serious shortand long-term complications, such as pancreatitis or coronary artery disease, it is critically important that health professionals recognize xanthomas when they occur, so that any underlying systemic disease can be identified and treated. The cutaneous xanthomas can be divided into 4 main subgroups based on their clinical appearance and location; these subgroups are the tendinous, planar, tuberous, and eruptive xanthomas. In this context, an important consideration is that certain types of xanthomas are more characteristic of specific underlying forms of hyperlipidemia than are other types.1,2 Tendinous xanthomas are usually caused by familial hypercholesterolemia or other disorders causing accumulation of low-density lipoproteins. These xanthomas appear as firm subcutaneous nodules found predominantly on the Achilles tendon and extensor tendons of digits. In contrast, planar xanthomas can occur at various sites. The most common planar xanthomas are xanthelasma palpebrarums, which occur on the eyelids and are suggestive of underlying hypercholesterolemia, especially in younger patients. However, xanthelasma palpebrarums also may occur in individuals without elevated cholesterol levels, and they also are associated, though less frequently, with other types of xanthomas. Planar xanthomas appearing on the palmar and finger creases are pathognomonic of familial dysbetalipoproteinemia type III. Since diffuse planar xanthomas symmetrically located on the neck and trunk are associated with paraproteinemia, hematologic malignancies should be considered in affected individuals. Tuberous xanthomas most commonly develop as coalescing dermal papules on elbows and knees, and are associated with various forms of hyperlipoproteinemia.2 Eruptive xanthomas are clinically characterized by clustered, yellowish papules with diameters of 2 to 5 mm (Figure). The lesions are usually asymptomatic, but they might itch while developing and might have an erythematous halo. They favor the extensor surfaces of arms and legs, buttocks, and trunk,1,2 and they also may develop in areas of injury,

ASSESSMENT Physical examination of the patient, who was extremely obese, revealed hundreds of yellowish, partially concave papules located on the extensor surfaces of her arms, legs, and trunk (Figure). The rash was most severe at mechanically stressed sites, such as elbows and knees, where the papules coalesced to form plaques. A few papules also were noted on the ear helices. The conjunctiva and sclera appeared normal. The clinical exam was otherwise unremarkable. Laboratory examination revealed chylomicronemia with triglyceride and total cholesterol levels of 8665 and 1560 mg/dL, respectively. The patient was hyperglycemic, with free glucose of 212 mg/dL. HbA1c was elevated at 10.6%, and urine protein was at 526 mg/L. Acute pancreatitis was excluded. Skin biopsy revealed an accumulation of foamy histiocytes in the upper dermis. There was no family history of inherited forms of hyperlipidemia.

Requests for reprints should be addressed to C.S. Seitz, MD, Department of Dermatology, Venerology and Allergology, University of Würzburg, Josef Schneider Strasse 2, D-97080 Würzburg, Germany. E-mail address: [email protected]

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2007.04.013

776

The American Journal of Medicine, Vol 120, No 9, September 2007

Figure

A rapidly developing rash of yellowish papules.

Trautmann et al

Pickled Plum of a Problem

infection, or mechanical irritation.3-6 Histologically, eruptive xanthomas represent an accumulation of lipid-filled macrophages, or “foam cells,” in the upper dermis. When plasma lipid concentrations are elevated, lipoproteins permeate the walls of dermal capillaries, depositing lipids in the dermis; there, they are phagocytised by macrophages.7 Although xanthomas are considered a sign of hyperlipidemia and correlate somewhat with serum lipid levels, most patients with lipoprotein abnormalities do not develop skin lesions.2 Eruptive xanthomas may be found in patients with primary hyperlipidemia due to familial lipoproteinemia types I, III, IV, and V (classification by Fredrickson), but the most common cause is secondary hypertriglyceridemia from uncontrolled diabetes mellitus.8-10 Two mechanisms are involved in the development of hypertriglyceridemia in this setting: defective lipolysis, which leads to impaired triglyceride clearance, and overproduction of very low-density lipoproteins. However, triglyceridemia varies widely in hyperglycemic patients, indicating that certain diabetic subjects are more prone to development of elevated lipid levels.11 Infrequently, development of xanthomas may be the initial manifestation of type-II diabetes, or they may occur in the early course of the disease, usually associated with elevated levels of fasting glucose, HbA1c, serum triglycerides, and total cholesterol.12,13 Other causes of secondary hypertriglyceridemia, such as hypothyroidism, cholestasis, and nephropathy also should be excluded in patients with eruptive xanthomas. Secondary hypertriglyceridemia also can be induced by several medications, including isotretinoin, indomethacin, and glucocorticoids, and it is relatively frequently induced by antipsychotic drugs such as clozapine and olanzapine.14

MANAGEMENT Usually, xanthomas vanish with normalization of plasma lipid levels, which may be achieved by dietary control, alcohol abstinence, and treatment with lipid-lowering agents, such as fibrates.5,10,13 In the present case, the umeboshi diet was discontinued, and the patient was put on a

777 strict diet with reduced carbohydrate and total fat intake. Her blood glucose was optimized by modifying her antidiabetic therapy. With this treatment regimen, the patient’s total triglyceride level dropped to 717 mg/dL after 7 days and subsequently normalized. The skin lesions slowly disappeared, and 5 months later, only discrete hyperpigmentation could be observed.

References 1. Hayes BB, Boyd AS. Eruptive xanthomas. Papules may indicate underlying lipid disorder. Postgrad Med. 2005;118:11-12. 2. Parker F. Xanthomas and hyperlipidemias. J Am Acad Dermatol. 1985;13:1-30. 3. Goldstein GD. The Koebner response with eruptive xanthomas. J Am Acad Dermatol. 1984;10:1064-1065. 4. Cowen EW, Rabel A, Nava V, et al. Eruptive linear papules on the abdomen of a woman with ascites. J Am Acad Dermatol. 2004;51: 75-78. 5. Scavo S, Magro G, Gurrera A, et al. Isomorphic response in eruptive xanthomas. Dermatology. 2004;209:66-68. 6. Barker DJ, Gould DJ. The Koebner phenomenon in eruptive xanthoma. Arch Dermatol. 1979;115:112. 7. Parker F, Bagdade JD, Odland GF, Bierman EL. Evidence for the chylomicron origin of lipids accumulating in diabetic eruptive xanthomas: A correlative lipid biochemical, histochemical, and electron microscopic study. J Clin Invest. 1970;49:2172-2187. 8. Cruz PD, East C, Bergstresser PR. Dermal, subcutaneous, and tendon xanthomas: diagnostic markers for specific lipoprotein disorders. J Am Acad Dermatol. 1988;19:95-111. 9. Archer CB, MacDonald DM. Eruptive xanthomata in type V hyperlipoproteinaemia associated with diabetes mellitus. Clin Exp Dermatol. 1984;9:312-316. 10. Nayak KR, Daly RG. Eruptive xanthomas associated with hypertriglyceridemia and new-onset diabetes mellitus. N Engl J Med. 2004; 350:1235. 11. Fulop M, Eder HA. Plasma triglycerides and cholesterol in diabetic ketosis. Arch Intern Med. 1989;149:1997-2002. 12. Shinozaki S, Itabashi N, Rokkaku K, et al. Diabetic lipemia with eruptive xanthomatosis in a lean young female with apolipoprotein E4/4. Diabetes Res Clin Pract. 2005;70:183-192. 13. Brunzell JD, Ayyobi AF. Dyslipidemia in the metabolic syndrome and type 2 diabetes mellitus. Am J Med. 2003;115:24S-28S. 14. Chang HY, Ridky TW, Kimball AB, et al. Eruptive xanthomas associated with olanzapine use. Arch Dermatol. 2003;139:1045-1048.

The American Journal of Medicine (2007) 120, 778-780

ECG IMAGE OF THE MONTH Julia H. Indik, MD, PhD, Section Editor

Can’t Catch a Breath Bishoy Faltas,a Ming-Yan Chowb a

Department of Internal Medicine and bPulmonary Medicine and Critical Care Unit, Rochester General Hospital, Rochester, NY

PRESENTATION An electrocardiogram (ECG) is always an appropriate part of the initial work-up for the severely dyspneic patient. In this case, the patient was a 72-year-old man with a history of coronary artery disease, diabetes, hypertension, chronic obstructive pulmonary disease, and chronic renal insufficiency. He complained of chest discomfort, diaphoresis, and progressive shortness of breath. He led a sedentary lifestyle and was a heavy smoker.

ASSESSMENT Upon physical examination, the patient was found to be in acute distress with profound shortness of breath, a respiratory rate of 30, and a heart rate of 130 bpm. His blood pressure was 110/68 mm Hg. He experienced 2 brief syncopal episodes while in the emergency room. His lungs were clear, and his left lower leg was swollen but not tender and did not have palpable veins. The patient’s chest X-ray did not show any acute changes from previous x-rays. Results of arterial blood-gas analysis (room air) revealed a pH of 7.34; PCO2, 23 mm Hg; PO2, 68 mm Hg; and bicarbonate, 12 mEq/L. His creatinine was 2.7 mg/dL, brain natriuretic peptide (BNP) was 312 pg/mL, and troponin was 0.76 ng/dL. An initial electrocardiogram (ECG) taken upon his arrival at the emergency room revealed a number of abnormalities (Figure 1), and no changes were apparent in a second ECG obtained 5 hours later. A bedside echocardiogram showed slight enlargement of the right ventricle and hypokinesia. The ECGs showed sinus tachycardia, S waves in lead I and Q waves in lead III (both ⬎ 1.5 mm), T-wave inversion in lead III, incomplete right bundle-branch block with rSR= pattern in V1, possible enlargement of the right ventricle, and inverted T waves in the right precordial leads (V1 to

Requests for reprints should be addressed to Bishoy Faltas, MD, Internal Medicine Residency Program, Rochester General Hospital, Rochester, NY 14621. E-mail adrress: [email protected]

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2007.05.002

V3). ST-segment depression also was present in leads I, aVL, and V4 to V6 (Figure 1).

DIAGNOSIS The patient’s ECG findings and clinical assessment raised a strong suspicion of pulmonary embolism (PE). Seventy percent of patients with PE have abnormal ECGs.1 Our patient’s ECG displayed many patterns commonly associated with PE, including sinus tachycardia, incomplete right bundle-branch block, an SIQIIITIII pattern suggestive of right ventricular strain, and T-wave inversions in leads V1 to V3.2 Although some PE patients also exhibit other abnormalities, our patient did not exhibit premature ventricular extrasystoles, atrial fibrillation, low-voltage QRS complexes, or pseudoinfarction simulating anterior or inferior wall myocardial infarction.2,4 Although the patient’s ECG results were consistent with PE, ECG findings alone are neither specific nor sensitive enough to confirm or exclude PE.5 Studies of the occurrence of specific ECG findings in patients with PE have yielded widely varying estimates, and some of those studies did not include control subjects without PE.3,5 The classically cited S1Q3T3 pattern was found in only about 12% of patients with angiographically documented PE.6 Perhaps of more significance is the confirmed higher prevalence of ECG abnormalities in patients with the most extensive embolization, as determined using imaging studies,1,7-9 making it unusual to find a normal ECG in the setting of massive PE.9 Nonetheless, the ECG remains an appropriate part of the initial workup because it can exclude other diagnoses such as myocardial infarction. When considered in the context of the patient’s clinical picture, it also might raise clinical suspicion and increase the pretest probability of acute PE.10 In the present case, the ECG contributed to our formulation of a high pretest probability of PE. Because the patient’s renal function was severely impaired, we did not order computed tomography (CT) pulmonary angiography. A ventilation/perfusion scan performed when the patient stabilized revealed multiple perfusion defects in the right upper

Faltas and Chow

Can’t Catch a Breath

779

Figure 1

ECG at the time of presentation.

and right lower lobes, with no matching ventilation defects, confirming the diagnosis of PE. The pathophysiology of ECG changes in PE might be explained by any of several different hypotheses, including hypoxia, coronary hypoperfusion, right-ventricular dilatation secondary to acute pulmonary hypertension, anatomic rotation, and autonomic nervous system dysfunction.11

Figure 2

MANAGEMENT The patient was rapidly transferred to the medical ICU. Because we felt that the probability of PE was high, based on his clinical and ECG findings, we started him on intravenous heparin therapy. He stabilized in a few hours and, as noted above, was sent for a ventilation/perfusion scan, which confirmed the PE diagnosis. He was subse-

ECG 1 week after admission.

780

The American Journal of Medicine, Vol 120, No 9, September 2007

quently maintained on intravenous heparin and, after a few days, he was discharged on coumadin. He continued to improve over time. Some studies have suggested that ECG abnormalities, such as the SIQIIITIII pattern, right bundle-branch block, and anterior T-wave inversion, can return to normal after PE is treated.8,12,13 An ECG obtained in our patient 1 week after treatment began showed normal sinus rhythm, resolution of S waves in lead I but persistence of Q waves in lead III, and persistent T-wave abnormalities (Figure 2). Unfortunately, we did not have the chance to obtain subsequent serial ECGs and thus were unable to observe complete resolution of these abnormalities.

References 1. Stein PD, Terrin ML, Hales CA, et al. Clinical, laboratory, roentgenographic, and electrocardiographic findings in patients with acute pulmonary embolism and no pre-existing cardiac or pulmonary disease. Chest. 1991;100:598-603. 2. Chan TC, Vilke GM, Pollack M, et al. Electrocardiographic manifestations: Pulmonary embolism. J Emerg Med. 2001;21:263-270. 3. Rodger M, Makropolous D, Turek M, et al. Diagnostic value of the electrocardiogram in suspected pulmonary embolism. Am J Cardiol. 2000;86:807-809.

4. Ulman E, Brady WJ, Perron AD, et al. Electrocardiographic manifestations of pulmonary embolism. Am J Emerg Med. 2001;19:514-519. 5. Brown G, Hogg K. Best evidence topic report. Diagnostic utility of electrocardiogram for diagnosing pulmonary embolism. Emerg Med J. 2005;22:729-730. 6. Sasahara AA, Hyers TM, Cole CM, et al. The urokinase pulmonary embolism trial: A national cooperative study. Circulation. 1973; 47(suppl 2):1-108. 7. Iles S, Le Heron CJ, Davies G, et al. ECG score predicts those with the greatest percentage of perfusion defects due to acute pulmonary thromboembolic disease. Chest. 2004;125:1651-1656. 8. Ferrari E, Imbert A, Chevalier T, et al. The ECG in pulmonary embolism: predictive value of negative T waves in precordial leads: 80 case reports. Chest. 1997;111:537-543. 9. Daniel KR, Courtney DM, Kline JA. Assessment of cardiac stress from massive pulmonary embolism with 12-lead ECG. Chest. 2001; 120:474-481. 10. Sinha N, Yalamanchili K, Sukhija R, et al. Role of the 12-lead electrocardiogram in diagnosing pulmonary embolism. Cardiol Rev. 2005;13:46-49. 11. Denton TA, Litvack F, Siegel RJ. Pseudoinfarction in patients with massive pulmonary embolism. West J Med. 1986;145:98-101. 12. McGinn S, White PD. Acute cor pulmonale resulting from pulmonary embolism. JAMA. 1935;104:1473-1480. 13. Durant TM, Ginsburg IVY, Roesler H, et al. Transient bundle branch block and other electrocardiographic changes in pulmonary embolism. Am Heart J. 1939;17:423-30.

The American Journal of Medicine (2007) 120, 781-782

IMAGES IN RADIOLOGY Michael A. Bettmann, MD, Section Editor

Narrowing in on a Diagnosis Mohit Jain,a Satish Gopal,a Leslie Trienb a

Department of Internal Medicine, bDepartment of Radiology, Norwalk Hospital, CT.

PRESENTATION Shortness of breath is common in cardiopulmonary disorders, but might also arise from many other disease processes, such as anemia, acidosis, or disruption in central nervous control of respiratory drive. In this case, the patient was a 46-year-old male with an otherwise unremarkable medical history who had been experiencing shortness of breath for 2 months. He reported that he also had been experiencing progressive swelling, and physical examination confirmed generalized edema. Shortly after admission, the patient developed respiratory failure requiring endotracheal intubation, and aggressive diuresis was initiated.

ASSESSMENT Surface echocardiography revealed a left ventricular ejection fraction of 45% with concentric left ventricular hypertrophy, moderate aortic insufficiency, and left atrial enlargement. After extubation, the patient was found to have a loud bruit in his back that radiated throughout the precordium. Four-limb blood-pressure measurements revealed isolated left-arm hypertension; his left arm was 200/100 mm Hg, whereas his right arm and the legs were 130/70 mm Hg. There was also pulse asymmetry, with a fuller radial pulse on the left than on the right.

DIAGNOSIS A computed tomography (CT) aortogram revealed narrowing of the aorta (aortic coarctation) distal to the origin of the left subclavian artery (Figures 1-3). It also revealed that the patient’s right subclavian artery originated at the site of coarctation rather than the brachiocephalic artery and curved behind the esophagus; this aberration accounted for the observed isolated left-arm hypertension. Extensive collateral circulation arising from the dilated proximal left subclavian artery also was demonstrated. Requests for reprints should be addressed to Satish Gopal, MD, MPH, Department of Internal Medicine, Norwalk Hospital, 34 Maple Street, Norwalk, CT 06856. E-mail address: [email protected]

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2007.06.001

Aortic coarctation accounts for 0.2% of cases of secondary hypertension1 and for 6% to 8% of all congenital heart defects. It occurs more than twice as frequently in males as in females. The mean survival of patients with untreated aortic coarctation is 34 years.2 Death can occur secondary to heart failure, aortic dissection or rupture, endocarditis, endarteritis, cerebral hemorrhage, ischemic heart disease, or concomitant aortic valve disease. The most common symptoms of aortic coarctation in adults are shortness of breath and fatigue.3 Asymptomatic hypertension is also common. Other clinical manifestations include headache, nosebleeds, heart failure, aortic dissection, and cramping pain in the lower extremities. Physical exam will reveal differences in the pulse amplitudes and blood pressures of the 4 extremities. The coarctation usually occurs distal to the origin of the left subclavian artery (postductal), giving rise to a pulse and pressure gradient between the arms and legs. Less commonly, narrowing occurs proximal to the origin of the left subclavian artery (preductal), causing a gradient between the right arm and the remaining extremities. Isolated left-arm hypertension is highly unusual; in this case, it was due to the aberrant right subclavian artery. Aortic coarctation commonly occurs together with other anomalies of the heart and aorta, especially bicuspid aortic valve, which occurs in 30% to 85% of patients with aortic coarctation.4,5 Other anomalies that might occur in conjunction with aortic coarctation include tubular hypoplasia of the aortic arch, ventricular septal defect, patent ductus arteriosus, left ventricular outflow or inflow tract obstruction, positional anomalies of the great vessels, including transposition, and mitral stenosis.4,6 An aberrant right subclavian artery that arises from a point distal to the origin of the left subclavian artery and passes behind the esophagus to the right arm occurs in 1% of the population overall.7 It has been rarely described together with aortic coarctation and was the only other cardiac anomaly identified in our patient. The noncardiac anomaly most commonly associated with aortic coarctation in patients is intracranial aneurysm, which occurs in up to 10% of aortic coarctation patients and may present as a life-threatening cerebral hemorrhage.8

782

The American Journal of Medicine, Vol 120, No 9, September 2007

Figure 1 Maximum intensity projection CT coronal image demonstrating aortic coarctation (arrow) separating the aortic arch (AA) and descending aorta (DA). An aberrant right subclavian artery (RSC) arising at the site of coarctation is also visible.

MANAGEMENT Aortic coarctation can be treated by surgical repair or endovascular therapy (balloon angioplasty with or without stenting), but controversy remains as to which of these therapeutic approaches is optimal.4 A recent review of endovascular therapy versus surgical repair demonstrated similar improvements in blood pressure control and morbidity,

Figure 3 Three-dimensional reconstructed CT sagittal image demonstrating aortic coarctation (arrow) separating the aortic arch (AA) and descending aorta (DA). A large, dilated left subclavian artery (LSC) is also shown.

but the incidence of restenosis and recurrence requiring reintervention was much higher with endovascular treatment.9 During his hospital stay, the patient responded to diuresis by losing weight (56 kg), and he exhibited clinical improvement. For treatment of his coarctation, he was referred to a clinic specializing in adult congenital heart disease, where catheter-based stenting of his coarctation was recommended as the least invasive therapeutic option. However, as he had no health insurance at the time, he declined all interventions. At present, social workers are assisting him in obtaining health insurance so that the delayed treatment can begin.

References

Figure 2 Maximum intensity projection CT sagittal image demonstrating aortic coarctation (arrow) separating the aortic arch (AA) and descending aorta (DA). The left common carotid artery (LCC) and a large, dilated left subclavian artery (LSC) are also visible.

1. Katira R, Rathore VS, Lip GY, Singh SP. Coarctation of aorta. J Hum Hypertens. 1997;11:537-538. 2. Campbell M. Natural history of coarctation of the aorta. Br Heart J. 1970;32:633-640. 3. Brouwer RM, Erasmus ME, Ebels T, Eijgelaar A. Influence of age on survival, late hypertension, and recoarctation in elective aortic coarctation repair. J Thorac Cardiovasc Surg. 1994;108:525-531. 4. Hanley FL. The various therapeutic approaches to aortic coarctation: Is it fair to compare? J Am Coll Cardiol. 1996;27:471-472. 5. Nihoyannopoulos P, Karas S, Sapsford RN, et al. Accuracy of twodimensional echocardiography in the diagnosis of aortic arch obstruction. J Am Coll Cardiol. 1987;10:1072-1077. 6. Levine JC, Sanders SP, Colan SD, et al. The risk of having additional obstructive lesions in neonatal coarctation of the aorta. Cardiol Young. 2001;11:44-53. 7. Bisognano JD, Young B, Brown JM, et al. Diverse presentation of aberrant origin of the right subclavian artery: two case reports. Chest. 1997;112:1693-1697. 8. Connolly HM, Huston J III, Brown RD Jr, et al. Intracranial aneurysms in patients with coarctation of the aorta: A prospective magnetic resonance angiographic study of 100 patients. Mayo Clin Proc. 2003;78:1491-1499. 9. Carr JA. The results of catheter-based therapy compared with surgical repair of adult aortic coarctation. J Am Coll Cardiol. 2006 21;47:1101-1107.

The American Journal of Medicine (2007) 120, 783-790

CLINICAL RESEARCH STUDY

Efficacy of Short-Course Antibiotic Regimens for Community-Acquired Pneumonia: A Meta-analysis Jonathan Z. Li, MD,a Lisa G. Winston, MD,a,b Dan H. Moore, PhD,c Stephen Bent, MDd a

Department of Medicine, bInfectious Diseases Division, cDepartment of Epidemiology and Biostatistics, and dGeneral Internal Medicine Section, San Francisco VA Medical Center, University of California, San Francisco. ABSTRACT PURPOSE: There is little consensus on the most appropriate duration of antibiotic treatment for communityacquired pneumonia. The goal of this study is to systematically review randomized controlled trials comparing short-course and extended-course antibiotic regimens for community-acquired pneumonia. METHODS: We searched MEDLINE, Embase, and CENTRAL, and reviewed reference lists from 1980 through June 2006. Studies were included if they were randomized controlled trials that compared short-course (7 days or less) versus extended-course (⬎7 days) antibiotic monotherapy for communityacquired pneumonia in adults. The primary outcome measure was failure to achieve clinical improvement. RESULTS: We found 15 randomized controlled trials matching our inclusion and exclusion criteria comprising 2796 total subjects. Short-course regimens primarily studied the use of azithromycin (n ⫽ 10), but trials examining beta-lactams (n ⫽ 2), fluoroquinolones (n ⫽ 2), and ketolides (n ⫽ 1) were found as well. Of the extended-course regimens, 3 studies utilized the same antibiotic, whereas 9 involved an antibiotic of the same class. Overall, there was no difference in the risk of clinical failure between the short-course and extended-course regimens (0.89, 95% confidence interval [CI], 0.78-1.02). In addition, there were no differences in the risk of mortality (0.81, 95% CI, 0.46-1.43) or bacteriologic eradication (1.11, 95% CI, 0.76-1.62). In subgroup analyses, there was a trend toward favorable clinical efficacy for the short-course regimens in all antibiotic classes (range of relative risk, 0.88-0.94). CONCLUSIONS: The available studies suggest that adults with mild to moderate community-acquired pneumonia can be safely and effectively treated with an antibiotic regimen of 7 days or less. Reduction in patient exposure to antibiotics may limit the increasing rates of antimicrobial drug resistance, decrease cost, and improve patient adherence and tolerability. © 2007 Elsevier Inc. All rights reserved. KEYWORDS: Antibiotics; Community-acquired pneumonia; Pneumonia; Short-course

Community-acquired pneumonia is one of the leading causes of morbidity and mortality in the world. In the United States, an estimated 2-3 million cases of communityacquired pneumonia occur annually, resulting in an estimated 10 million physician visits and 600,000 hospitalizations, with a total annual cost of over $20 billion.1-3 The most commonly isolated pathogen is Streptococcus pneuThe data in this article were presented in part at the 46th Annual Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco, CA, September 27-30, 2006 (Session 162, Paper L-1458). Requests for reprints should be addressed to Jonathan Z. Li, MD, Department of Medicine, University of California, San Francisco, Box 0862, 5H22, San Francisco, CA 94143-0862. E-mail address: [email protected]

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2007.04.023

moniae, especially in bacteremic and hospitalized patients. Other common causes of community-acquired pneumonia include Haemophilus influenzae and the “atypical” pathogens, which include Mycoplasma pneumoniae, Chlamydophila pneumoniae, and Legionella pneumophilia. The atypical organisms cannot be differentiated from other etiologies on the basis of clinical symptoms or radiographic appearance and are infrequently isolated in clinical practice. Overall, the bacteriologic etiology of community-acquired pneumonia is undetermined in 40%-60% of clinical studies and in the vast majority of cases in actual practice.1,2,4-7 Without adequate clinical trials, the empiric treatment of community-acquired pneumonia continues to be challenging, with a myriad of recommendations on the length of

784

The American Journal of Medicine, Vol 120, No 9, September 2007

treatment and types of antibiotics.1,3,8-12 The drawbacks of viewers (JL and LW) independently evaluated full text antibiotic overuse are becoming increasingly apparent and articles to determine eligibility for inclusion into the study include growing antibiotic resistance, rising costs, and poand independently extracted data from relevant trials. Distentially severe side effects such as Clostridium difficile crepancies between the reviewers were resolved through infections. Recent studies have already begun to question discussion. the need for additional antimicrobial agents to cover atypical orOutcomes ganisms in community-acquired The primary outcome measure CLINICAL SIGNIFICANCE pneumonia.6,13 Another method to was failure to achieve clinical imdecrease antibiotic usage is to deprovement or cure. Secondary out● Adults with mild-moderate communitycrease the length of antibiotic comes were mortality, bacterioacquired pneumonia can be effectively treatment. Currently, a range of logic failure, and adverse events. treated with an antibiotic regimen of 7 recommendations can be found reThe outcome measures and time days or less. garding the duration of treatment, to outcome assessment were demost encompassing a treatment ● This result is consistent among the 4 anfined by each individual study. course of 5-14 days.1,7-9,14 Several tibiotic classes studied (macrolide, fluoroClinical failure was determined by recent studies have suggested the the investigators of each study quinolone, beta-lactam, and ketolide). clinical effectiveness and benegenerally based on clinical symp● There is a trend toward decreased adfits of shorter duration antimicrotoms and the need for additional verse events with antibiotic regimens of bial therapy for lower-respiratory antibiotics. The time to outcome 7 days or less. tract infections and communityassessment was generally between acquired pneumonia.15-20 In order 10 and 42 days. The quality of to further define the appropriate each study was evaluated using length of antibiotic treatment for the Jadad scoring system as previcommunity-acquired pneumonia, we performed a metaously described, which takes into account allocation generanalysis of randomized-controlled trials comparing shortation, concealment, and dropouts.21 course (ⱕ7-day) versus extended-course (⬎7-day) antibiotic regimens.

Statistical Analysis

METHODS Search Strategy We used the Cochrane Central Register of Controlled Trials, Medline, and Embase to find publications from 1980 through June 2006. In the Cochrane database, the record title was searched for the keyword “pneumonia.” Articles in Medline and Embase were found by searching for clinical studies or trials with the word “pneumonia” in the title and without the following keywords in the title: “pneumocystis,” “aspiration,” “aspirate,” “nosocomial,” “ventilator,” “ventilation,” “ventilated,” “pediatric,” “paediatric,” “child,” “childhood,” and “children.” In addition, we reviewed the reference lists from review articles and the identified clinical trials, as well as medication inserts and drug manufacturer websites to identify other relevant studies. No language restrictions were applied during the search process.

For the primary data analysis, we used the intention-to-treat or modified intention-to-treat (participants who were randomized and received at least one dose of study medication) principle. In studies that only reported a per-protocol analysis, we conservatively assumed that all dropouts were treatment failures for inclusion in the intention-to-treat analysis. Clinical success using the per-protocol patient population also was evaluated. Stata version 9.0 (StataCorp LP, College Station, Tex) was employed to calculate the relative risks (RR) and 95% confidence intervals (CI) using the fixed-effects model (a random-effects model also was used and was found to not significantly alter the results). To test for publication bias, we used the Stata program metabias, which performs the Begg’s and Egger’s regression asymmetry test for publication bias. To test for heterogeneity, we calculated a chi-squared statistic that is the sum of the squared differences between each study and the summary mean divided by the variance (StataCorp LP).

Study Selection Our inclusion criteria required studies to be randomized controlled trials in adults (age ⱖ12 years) that compared the clinical efficacy of a short-course (7 days or less) antibiotic monotherapy regimen versus an extended course (⬎7 days) regimen. All patients had radiographically confirmed pneumonia. Noncomparative and nonrandomized studies were excluded, as were trials with a large proportion of patients with bronchitis, chronic obstructive pulmonary disease exacerbations, or health-care-associated pneumonias. Two re-

RESULTS Description of Studies Our search strategy identified more than 3700 potential references (Figure 1). Most studies were excluded due to the fact that they were not randomized controlled trials or did not compare a short-course versus an extended-course antibiotic regimen. Fifteen studies met our inclusion and exclusion criteria and were included in this meta-analysis

Li et al

Short-Course Antibiotics for CAP

Figure 1

785

Study flow diagram.

(Table 1). These studies were published between 1990 and 2004 and comprised a total of 2796 subjects. Ten studies enrolled more than 100 participants (range, 42-528 subjects). The mean age of the participants ranged from 40 to 64 years. All of the trials included only patients with mildmoderate community-acquired pneumonia. Eight of the studies exclusively used oral antibiotics and excluded those with more severe disease requiring intravenous antibiotics. The other trials excluded patients with signs of clinical decompensation (eg, patients requiring intensive care unit stay, Pneumonia Severity Index ⬎130).22 In the 10 studies examining the efficacy of short-course azithromycin,23-32 6 trials used 3-day regimens,26-30,32 whereas 4 trials used 5-day regimens.23-25,31 Most compared it with another mac-

Table 1

rolide, although the comparative drug was Cefaclor in one study25 and in another, multiple comparative antibiotics were used.28 Other short-course regimens studied included fluoroquinolones (n ⫽ 2),33,34 beta-lactams (n ⫽ 2),35,36 and a ketolide (n ⫽ 1).37 The majority of extended-course antibiotic regimens involved macrolide antibiotics (n ⫽ 9), but also examined the beta-lactam (n ⫽ 4) and fluoroquinolone (n ⫽ 1) antibiotic classes. Although the majority of studies examined short-course macrolide antibiotics, the distribution of subjects was more evenly distributed among the different antibiotic classes. Of the 2796 total participants, 39% were in studies examining short courses of macrolide antibiotics, 30% in the fluoroquinolone trials, 20% in the ketolide study, and 11% in the studies examining short-course beta-lactam antibiotics. Both inpatients and outpatients were represented, with 2 studies performed exclusively in outpatients,27,28 4 studies with only inpatients,23,29,30,35 and 6 studies evaluating both inpatients and outpatients26,31-34,37 (2 studies did not specify).24,25 All studies identified participants by a combination of clinical symptoms and radiographic features. Several studies had more targeted inclusion criteria. Bohte et al divided their cohort into patients suspected to have pneumococcal and nonpneumococcal pneumonia based on clinical and microbiological criteria.23 The pneumococcal subgroup was not included in this analysis, as the duration of antibiotic use was not specified for the comparison medication. In 2 studies, both involving azithromycin compared with another macrolide, the patient population was restricted to those deemed to have “atypical pneumonia” by either clinical/ radiographic criteria or through serologic antibody titers.30,31 Finally, in the study by Leophonte et al comparing gemifloxacin and amoxicillin/clavulanic acid, the participants were limited to those who were thought to have likely

Characteristics of Included Studies

Study

Short-Course

Extended-Course

n

Mean Age*

Bohte et al, 199523 Brion et al, 199024 Dunbar et al, 200333

Azithromycin, 5 d Azithromycin, 5 d Levofloxacin, 5 d

Erythromycin, 10 d Josamycin, 10 d Levofloxacin, 10 d

42 97 528

61 53 54

Kinasewitz & Wood, 199125 Kobayashi et al, 199526 Leophonte et al, 200434 Leophonte et al, 200235 O’Doherty & Muller, 199827 Rahav et al, 200428 Rizzato et al, 199529 Schonwald et al, 199430 Schonwald et al, 199031 Siegel et al, 199936 Sopena et al, 200432 Tellier et al, 200437

Azithromycin, 5 d Azithromycin, 3 d Gemifloxacin, 7 d Ceftriaxone, 5 d Azithromycin, 3 d Azithromycin, 3 d Azithromycin, 3 d Azithromycin, 3 d Azithromycin, 5 d Cefuroxime, 7 d Azithromycin, 3 d Telithromycin, 5 or 7 d

Cefaclor, 10 d Clarithromycin, 14 d Amoxicillin/clav, 10 d Ceftriaxone, 10 d Clarithromycin, 10 d Multiple abx, 10 d Clarithromycin, 10 d Roxithromycin, 10 d Erythromycin, 10 d Cefuroxime, 10 d Clarithromycin, 10 d Clarithromycin, 10 d

119 163 320 244 203 108 40 150 101 52 70 559

42 Not reported 54 64 51 50 46 40 Not reported 61 43 42

*Mean age (years) is estimated to be the average age of the 2 arms if reported separately.

Time to Outcome Assessment Within 21 days of discharge 30 days 7-14 days after last dose of antibiotic 10-13 days 14 days 24-30 days 10 days 12-16 days 10-14 days 30 days 14 days 15-21 days 42 days 25-30 days 17-21 days

786

The American Journal of Medicine, Vol 120, No 9, September 2007

Table 2 Number of Patients with Community-Acquired Pneumonia Failing to Improve Clinically by Intention-to-Treat (ITT) and Per-Protocol (PP) Analysis ITT Analysis n/N Study 23

Bohte et al, 1995 Brion et al, 199024 Dunbar et al, 200333 Kinasewitz & Wood, 199125 Kobayashi et al, 199526 Leophonte et al, 200434 Leophonte et al, 200235 O’Doherty & Muller, 199827 Rahav et al, 200428 Rizzato et al, 199529 Schonwald et al, 199430 Schonwald et al, 199031 Siegel et al, 199936 Sopena et al, 200432 Tellier et al, 200437 Summary RR

PP Analysis n/N

Risk Ratios

Short-Course Extended-Course Short-Course Extended-Course ITT (95% CI)

PP (95% CI)

5/20 13/50 73/256 23/53 23/81 38/167 32/125 18/101 1/62 1/20 2/90 18/57 6/27 6/34 67/378

0.88 1.68 0.86 6.06 0.18 0.91 1.04 1.25 0.12 0.47 0.20

(0.28-2.82) (0.61-4.62) (0.44-1.66) (0.30-121.9) (0.02-1.43) (0.45-1.85) (0.56-1.93) (0.35-4.50) (0.02-0.99) (0.05-4.82) (0.02-1.86)

1.37 0.77 1.33 0.94

(0.25-7.47) (0.19-3.18) (0.71-2.49) (0.72-1.22)

6/22 9/47 97/272 27/66 25/82 32/153 34/119 18/102 6/46 3/20 10/60 12/44 5/25 8/36 34/181

4/19 9/46 15/198 2/32 1/59 13/115 17/94 5/88 1/62 1/20 1/89 0/39 3/24 3/31 35/320

5/21 5/43 17/192 0/39 6/63 14/113 16/92 4/88 6/46 2/19 3/53 0/32 2/22 4/32 12/146

0.92 1.36 0.80 1.06 0.93 1.09 0.90 1.01 0.12 0.33 0.13 1.16 1.11 0.79 0.94 0.89

(0.33-2.54) (0.64-2.88) (0.62-1.03) (0.70-1.62) (0.58-1.50) (0.72-1.65) (0.59-1.35) (0.56-1.83) (0.02-0.99) (0.04-2.94) (0.03-0.59) (0.63-2.14) (0.39-3.19) (0.31-2.05) (0.65-1.37) (0.78-1.02)

*Relative risk unable to be calculated for the PP population for Schonwald et al. due to the lack of patients who failed to improve in both arms of the study.

pneumococcal pneumonia based on clinical findings or examination of a respiratory sample.34

Risk of Clinical Failure The risk of clinical failure was the primary outcome measure for all studies. No significant differences were found between the risk of clinical failure in the short-course and extended-course arms (Table 2, Figure 2) either by intention-to-treat (RR 0.89, 95% CI, 0.78-1.02) or by per-protocol analysis (RR 0.94, 95% CI, 0.72-1.22). No significant

Figure 2

differences in outcome were found in those participants taking short-course macrolides (10 studies: RR 0.88, 95% CI, 0.71-1.09), fluoroquinolones (2 studies: RR 0.88, 95% CI, 0.71-1.08), or beta-lactam antibiotics (2 studies: RR 0.92, 95% CI, 0.63-1.36). A subgroup analysis of studies using 3-day regimens of azithromycin showed a significant reduction in clinical failures with the fixed-effects model (6 studies: RR 0.70, 95% CI, 0.51-0.96), but not the randomeffects model (6 studies: RR0.61, 95% CI, 0.34-1.10). The reported rate of clinical failure by per-protocol analysis was

Relative risk of clinical failure with short-course versus extended course antibiotic regimens.

Li et al

Short-Course Antibiotics for CAP

787

Figure 3 Relative risk of mortality with short-course versus extended-course antibiotic regimens. (The relative risk of mortality could not be calculated in 7 studies due to the lack of deaths in both arms).

8.9% for participants in the short-course antibiotic arm and 9.6% in the extended-course antibiotic arm.

Secondary Outcome Measures The overall mortality rate was 1.7%, with 7 studies reporting no deaths. Among those studies with at least one death, the mortality rates ranged from 0.9% to 6.7%. Among the analyzable studies, no significant differences were found in the risk of mortality between the 2 arms (Figure 3; RR 0.81, 95% CI, 0.46-1.43). Bacteriologic failure was reported in 7 studies and was generally defined as persistently positive cultures or an absence of culture/serologic testing results in those who had clinical failure. Prolonged antibiotic course was not associated with significantly improved bacteriologic response (RR 1.09, 95% CI, 0.75-1.58). Adverse events were defined as clinical symptoms or laboratory abnormalities deemed likely to be related to medication use by the investigators. There was a wide range in the reported rates of adverse events (2.3%-22.4%), with a mean of 14.1%. No significant differences in the risk of adverse events were found between the arms (RR 0.86, 95% CI, 0.71-1.04). No heterogeneity was found for the intentionto-treat analysis of clinical failure (P ⫽ .36), mortality (P ⫽ 1.0), bacteriologic eradication (P ⫽ .60), or for any subgroup analysis (P ⬎.1 for all comparisons). Tests for publication bias found no evidence of publication bias for any of the summary measures (P ⬎.1 for all comparisons). Unless noted above, analysis by the random effects model resulted in no significant differences among the results.

Methodologic Quality of Studies The quality of the studies was evaluated on the basis of adequate allocation randomization, concealment, and the

reporting of dropouts. This was done through the Jadad score as previously described.21 Eight studies were found to be of relatively high quality (Jadad score ⱖ3). A subgroup analysis including only the high quality studies found no significant differences in the risk of clinical failure with the use of short-course antibiotic regimens (RR 0.92, 95% CI, 0.80-1.07).

DISCUSSION In this meta-analysis, we found no significant differences between short-course and extended-course antibiotic regimens for the treatment of mild to moderate communityacquired pneumonia with respect to clinical success, mortality, bacteriologic success, and adverse events. The results were consistent across a wide range of analyses, including both the intention-to-treat and per-protocol patient populations, high-quality studies, and individual antibiotic classes. Both outpatients and inpatients were included, and 4 of the antibiotic classes most commonly used for community-acquired pneumonia (macrolide, fluoroquinolone, beta-lactam, and ketolide) were represented in this meta-analysis. In addition, the studies taken together included subjects with a wide mean age range. The optimal length of treatment for community-acquired pneumonia has been unclear, with current guidelines suggesting a regimen that varies from 5 to 14 days.5,7,8,14,38,39 In addition, the most recent Infectious Diseases Society of America and American Thoracic Society guidelines specifically recommend treatment until 72 hours after the patient becomes afebrile and until clinically stable.1,8,9 Several arguments have already been put forth for the use of shorter duration of treatment for pneumonia. Studies in children have demonstrated the equivalent efficacy of 3 days versus

788

The American Journal of Medicine, Vol 120, No 9, September 2007

5 days of antibiotic treatment for pneumonia.40,41 An older study of community-acquired pneumonia patients in Nigeria suggested that patients could be successfully treated with as little as 2.5 days of antibiotic therapy.42 In the treatment of ventilator-associated pneumonia, 8 days of antibiotic treatment was found to be as efficacious as 15 days of treatment in most cases.43 In addition, a study of patients with nosocomial pneumonia found, using serial bronchoscopy, that in only 6% of cases were initially isolated microbes not eradicated within just 3 days of treatment.44 One reason that azithromycin has been so frequently studied in short-course regimens is its pharmacokinetic properties that allow the drug to have high tissue concentrations for 3-4 days after completion of therapy.15,45 It is interesting to note that in 6 of 10 azithromycin trials included here, azithromycin was given for only 3 days, which would suggest that 1 week of an antibiotic without this type of prolonged activity should be sufficient. Finally, a recent multi-centered clinical trial in the Netherlands found that two thirds of patients with community-acquired pneumonia had improved clinically after just 3 days of treatment with amoxicillin. In those patients who had substantially improved, there was no benefit in taking additional antibiotics.46 The avoidance of extended-course antibiotic regimens may have many important benefits. Increasing rates of antimicrobial resistance has become a major concern for S. pneumoniae and other organisms causing community-acquired pneumonia.4,38,47,48 It is clear that one of the major causes is the frequency and length of antibiotic use, and subsequent selective pressures for resistance.16,43,49,50 The use of shorter course antibiotic regimens may help limit the spread of drug-resistant bacteria. Patient compliance is another factor to consider, as several studies have shown improved patient adherence with regimens of ⬍7 days compared with longer courses.50-52 Finally, shorter courses of antibiotics can potentially reduce the risk of medication side effects. In this study, there was a trend toward lower adverse events with the short-course regimen. The issue of antibiotic resistance has become a grave concern, especially with the well-documented increase in macrolide-resistant S. pneumoniae.8,53 S. pneumoniae resistance to fluoroquinolones also has been documented worldwide, but resistance to respiratory fluoroquinolones (levofloxacin, moxifloxacin, and gemifloxacin) is still relatively rare in the United States.53-57 Although there was a trend toward favorable clinical efficacy for the short-course regimen in all antibiotic classes, many of the included studies are of relatively small size and little information can be extrapolated as to the effect of antimicrobial resistance. This study is intended to address whether a short duration of antibiotic therapy is adequate using the antimicrobial classes that have been studied, but the appropriate selection of the type of antibiotic for community-acquired pneumonia may evolve as resistance patterns change. At this time, macrolides are still recommended as first-line therapy for the outpatient treatment of community-acquired pneumonia

in patients who are previously healthy and have no risk factors for drug-resistant S. pneumoniae.8 It also is important to note that telithromycin has been linked to a number of cases of hepatotoxicity and caution should be advised when prescribing this antibiotic.58,59 One important limitation of this study is the underrepresentation of some classes of antibiotics. For example, no study of doxycycline was found that matched the inclusion criteria. In addition, only 1 ketolide, 2 beta-lactam, and 2 fluoroquinolone studies were found. However, because these studies were larger, the overall number of subjects receiving each class of antibiotic was more evenly distributed. Although 10 of 15 studies examined the efficacy of azithromycin, these studies involved only 39% of the total number of participants. The 2 fluoroquinolone trials enrolled 848 participants, or 30% of the total number, whereas the one ketolide study alone had 559 participants. Another limitation is that most trials included only mild-moderate pneumonia, with elderly patients generally under-represented in the study populations. Even in the inpatient studies, respiratory failure and septic shock were common exclusion criteria. Therefore, although the results of this metaanalysis should be generalizable to most adults, they cannot be extrapolated to those with severe community-acquired pneumonia. Finally, as with all recommendations, individual response to treatment should be taken into account. As discussed in the recent study by el Moussaoui et al,46 short courses of treatment are probably most appropriate in patients who have significantly improved with initial therapy. In summary, the available data suggest that adults with mild-moderate community-acquired pneumonia can be safely and effectively treated with an antibiotic regimen of 7 days or less. Given the potential cost savings and implications in reducing antimicrobial drug resistance, larger studies should be performed confirming these results across antibiotic classes.

ACKNOWLEDGMENTS We thank Gloria Won for her invaluable advice on the Embase and Medline literature searches. We are grateful for the help Yoshio Hall and Delphine Tuot provided in translating articles.

References 1. Bartlett JG, Dowell SF, Mandell LA, et al. Practice guidelines for the management of community-acquired pneumonia in adults. Infectious Diseases Society of America. Clin Infect Dis. 2000;31(2):347-382. 2. Mandell LA. Epidemiology and etiology of community-acquired pneumonia. Infect Dis Clin North Am. Dec 2004;18(4):761-776, vii. 3. File TM Jr, Garau J, Blasi F, et al. Guidelines for empiric antimicrobial prescribing in community-acquired pneumonia. Chest. 2004;125(5): 1888-1901. 4. Segreti J, House HR, Siegel RE. Principles of antibiotic treatment of community-acquired pneumonia in the outpatient setting. Am J Med. 2005;118(7 Suppl):21S-28S. 5. File TM Jr. Community-acquired pneumonia. Lancet. 2003;362(9400): 1991-2001.

Li et al

Short-Course Antibiotics for CAP

6. Mills GD, Oehley MR, Arrol B. Effectiveness of beta lactam antibiotics compared with antibiotics active against atypical pathogens in non-severe community acquired pneumonia: meta-analysis. BMJ. 2005;330(7489):456. 7. Niederman MS, Mandell LA, Anzueto A, et al. Guidelines for the management of adults with community-acquired pneumonia. Diagnosis, assessment of severity, antimicrobial therapy, and prevention. Am J Respir Crit Care Med. 2001;163(7):1730-1754. 8. Mandell LA, Wunderink RG, Anzueto A, et al. Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis. 2007;44(Suppl 2):S27-S72. 9. Mandell LA, Bartlett JG, Dowell SF, et al. Update of practice guidelines for the management of community-acquired pneumonia in immunocompetent adults. Clin Infect Dis. 2003;37(11):1405-1433. 10. Mandell LA, Marrie TJ, Grossman RF, et al. Canadian guidelines for the initial management of community-acquired pneumonia: an evidence-based update by the Canadian Infectious Diseases Society and the Canadian Thoracic Society. Clin Infect Dis. 2000;31(2):383-421. 11. Halm EA, Teirstein AS. Clinical practice. Management of communityacquired pneumonia. N Engl J Med. 2002;347(25):2039-2045. 12. Bjerre LM, Verheij TJ, Kochen MM. Antibiotics for community acquired pneumonia in adult outpatients. Cochrane Database Syst Rev. 2004(2):CD002109. 13. Shefet D, Robenshtok E, Paul M, Leibovici L. Empirical atypical coverage for inpatients with community-acquired pneumonia: systematic review of randomized controlled trials. Arch Intern Med. 2005; 165(17):1992-2000. 14. BTS guidelines for the management of community acquired pneumonia in adults. Thorax. 2001;56(Suppl 4):IV1-IV64. 15. Contopoulos-Ioannidis DG, Ioannidis JP, Chew P, Lau J. Meta-analysis of randomized controlled trials on the comparative efficacy and safety of azithromycin against other antibiotics for lower respiratory tract infections. J Antimicrob Chemother. 2001;48(5):691-703. 16. File TM Jr. Clinical efficacy of newer agents in short-duration therapy for community-acquired pneumonia. Clin Infect Dis. 2004;39(Suppl 3):S159-S164. 17. Hopkins S, Williams D. Five-day azithromycin in the treatment of patients with community-acquired pneumonia. Curr Ther Res Clin Exp. 1995;56(9):915-925. 18. Socan M. Treatment of atypical pneumonia with azithromycin: comparison of a 5-day and a 3-day course. J Chemother. 1998;10(1):64-68. 19. Mandell LA, File TM Jr. Short-course treatment of community-acquired pneumonia. Clin Infect Dis. 2003;37(6):761-763. 20. Kolditz M, Halank M, Hoffken G. Short-course antimicrobial therapy for community-acquired pneumonia. Treat Respir Med. 2005;4(4): 231-239. 21. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials. 1996;17(1):1-12. 22. Fine MJ, Auble TE, Yealy DM, et al. A prediction rule to identify low-risk patients with community-acquired pneumonia. N Engl J Med. 1997;336(4):243-250. 23. Bohte R, Van’t Wout JW, Lobatto S, et al. Efficacy and safety of azithromycin versus benzylpenicillin or erythromycin in communityacquired pneumonia. Eur J Clin Microbiol Infect Dis. 1995;14(3):182187. 24. Brion JP, Sedallian A, Le Noc P, et al. Azithromycin versus josamycin: treatment of 89 acute pneumonia cases. Pathol Biol (Paris). 1990;38(5 Pt 2):521-525. 25. Kinasewitz G, Wood RG. Azithromycin versus cefaclor in the treatment of acute bacterial pneumonia. Eur J Clin Microbiol Infect Dis. 1991;10(10):872-877. 26. Kobayashi H, Sakayori S, Koike T, et al. Clarithromycin-controlled randomized double-blind studies of azithromycin for treatment of pneumonia. Jpn J Chemother. 1995;43(8):757-774. 27. O’Doherty B, Muller O. Randomized, multicentre study of the efficacy and tolerance of azithromycin versus clarithromycin in the treatment of

789

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38. 39. 40.

41.

42. 43.

44.

45.

46.

47.

adults with mild to moderate community-acquired pneumonia. Eur J Clin Microbiol Infect Dis. 1998;17(12):828-833. Rahav G, Fidel J, Gibor Y, Shapiro M. Azithromycin versus comparative therapy for the treatment of community acquired pneumonia. Int J Antimicrob Agents. 2004;24(2):181-184. Rizzato G, Montemurro L, Fraioli P, et al. Efficacy of a three day course of azithromycin in moderately severe community-acquired pneumonia. Eur Respir J. 1995;8(3):398-402. Schonwald S, Barsic B, Klinar I, Gunjaca M. Three-day azithromycin compared with ten-day roxithromycin treatment of atypical pneumonia. Scand J Infect Dis. 1994;26(6):706-710. Schonwald S, Gunjaca M, Kolacny-Babic L, et al. Comparison of azithromycin and erythromycin in the treatment of atypical pneumonias. J Antimicrob Chemother. 1990;25(Suppl A):123-126. Sopena N, Martinez-Vazquez C, Rodriguez-Suarez JR, et al. Comparative study of the efficacy and tolerance of azithromycin versus clarithromycin in the treatment of community-acquired pneumonia in adults. J Chemother. 2004;16(1):102-103. Dunbar LM, Wunderink RG, Habib MP, et al. High-dose, short-course levofloxacin for community-acquired pneumonia: A new treatment paradigm. Clin Infect Dis. 2003;37(6):752-760. Leophonte P, File T, Feldman C. Gemifloxacin once daily for 7 days compared to amoxicillin/clavulanic acid thrice daily for 10 days for the treatment of community-acquired pneumonia of suspected pneumococcal origin. Respir Med. 2004;98(8):708-720. Leophonte P, Choutet P, Gaillat J, et al. Efficacy of a ten day course of ceftriaxone compared to a shortened five day course in the treatment of community-acquired pneumonia in hospitalized adults with risk factors. Med Mal Infect. 2002;32(7):369-381. Siegel RE, Alicea M, Lee A, Blaiklock R. Comparison of 7 versus 10 days of antibiotic therapy for hospitalized patients with uncomplicated community-acquired pneumonia: a prospective, randomized, doubleblind study. Am J Ther. 1999;6(4):217-222. Tellier G, Niederman MS, Nusrat R, et al. Clinical and bacteriological efficacy and safety of 5 and 7 day regimens of telithromycin once daily compared with a 10 day regimen of clarithromycin twice daily in patients with mild to moderate community-acquired pneumonia. J Antimicrob Chemother. 2004;54(2):515-523. File TM, Jr, Niederman MS. Antimicrobial therapy of communityacquired pneumonia. Infect Dis Clin North Am. 2004;18(4):993-1016. Restrepo MI, Anzueto A. Antimicrobial treatment of communityacquired pneumonia. Clin Chest Med. 2005;26(1):65-73. Pakistan Multicentre Amoxycillin Short Course Therapy (MASCOT) pneumonia study group. Clinical efficacy of 3 days versus 5 days of oral amoxicillin for treatment of childhood pneumonia: a multicentre double-blind trial. Lancet. 2002;360(9336):835-841. Agarwal G, Awasthi S, Kabra SK, et al. Three day versus five day treatment with amoxicillin for non-severe pneumonia in young children: a multicentre randomised controlled trial. BMJ. 2004;328(7443): 791. Awunor-Renner C. Length of antibiotic therapy in in-patients with primary pneumonias. Ann Trop Med Parasitol. 1979;73(3):235-240. Chastre J, Wolff M, Fagon JY, et al. Comparison of 8 vs 15 days of antibiotic therapy for ventilator-associated pneumonia in adults: a randomized trial. JAMA. 2003;290(19):2588-2598. Montravers P, Fagon JY, Chastre J, et al. Follow-up protected specimen brushes to assess treatment in nosocomial pneumonia. Am Rev Respir Dis. 1993;147(1):38-44. Foulds G, Shepard RM, Johnson RB. The pharmacokinetics of azithromycin in human serum and tissues. J Antimicrob Chemother. 1990; 25(Suppl A):73-82. el Moussaoui R, de Borgie CA, van den Broek P, et al. Effectiveness of discontinuing antibiotic treatment after three days versus eight days in mild to moderate-severe community acquired pneumonia: randomised, double blind study. BMJ. 2006;332(7554):1355. Richter SS, Heilmann KP, Beekmann SE, et al. The molecular epidemiology of Streptococcus pneumoniae with quinolone resistance mutations. Clin Infect Dis. 2005;40(2):225-235.

790

The American Journal of Medicine, Vol 120, No 9, September 2007

48. Karchmer AW. Increased antibiotic resistance in respiratory tract pathogens: PROTEKT US—an update. Clin Infect Dis. 2004;39(Suppl 3):S142-S150. 49. Doern GV. Antimicrobial use and the emergence of antimicrobial resistance with Streptococcus pneumoniae in the United States. Clin Infect Dis. 2001;33(Suppl 3):S187-S192. 50. Schrag SJ, Pena C, Fernandez J, et al. Effect of short-course, high-dose amoxicillin therapy on resistant pneumococcal carriage: a randomized trial. JAMA. 2001;286(1):49-56. 51. Kardas P. Patient compliance with antibiotic treatment for respiratory tract infections. J Antimicrob Chemother. 2002;49(6):897-903. 52. Reyes H, Guiscafre H, Munoz O, et al. Antibiotic noncompliance and waste in upper respiratory infections and acute diarrhea. J Clin Epidemiol. 1997;50(11):1297-1304. 53. Cunha BA. Antimicrobial therapy of multidrug-resistant Streptococcus pneumoniae, vancomycin-resistant enterococci, and methicillin-resistant Staphylococcus aureus. Med Clin North Am. 2006;90(6):11651182.

54. Karlowsky JA, Thornsberry C, Jones ME, et al. Factors associated with relative rates of antimicrobial resistance among Streptococcus pneumoniae in the United States: results from the TRUST Surveillance Program (1998-2002). Clin Infect Dis. 2003;36(8):963-970. 55. Davidson R, Cavalcanti R, Brunton JL, et al. Resistance to levofloxacin and failure of treatment of pneumococcal pneumonia. N Engl J Med. 2002;346(10):747-750. 56. Whitney CG, Farley MM, Hadler J, et al. Increasing prevalence of multidrug-resistant Streptococcus pneumoniae in the United States. N Engl J Med. 2000;343(26 I):1917-1924. 57. Powis J, McGeer A, Green K, et al. In vitro antimicrobial susceptibilities of Streptococcus pneumoniae clinical isolates obtained in Canada in 2002. Antimicrob Agents Chemother. 2004;48(9):3305-3311. 58. Clay KD, Hanson JS, Pope SD, et al. Brief communication: severe hepatotoxicity of telithromycin: three case reports and literature review. Ann Intern Med. 2006;144(6):415-420. 59. Ross D. The FDA and the case of Ketek. N Engl J Med. 2007;356(16): 1601-1604.

The American Journal of Medicine (2007) 120, 791-798

CLINICAL RESEARCH STUDY

The Safety of Oral Antifungal Treatments for Superficial Dermatophytosis and Onychomycosis: A Meta-analysis Chia-Hsuin Chang, MD, MSc,a,b Yinong Young-Xu, ScD, MS, MA,c Tobias Kurth, MD, ScD,b,d John E. Orav, PhD,b Arnold K. Chan, MD, ScDb,e a

Department of Internal Medicine, National Taiwan University Hospital, Taipei, Taiwan; bDepartment of Epidemiology, Harvard School of Public Health, Boston, Mass; cEpiPatterns, Haverhill, NH; dDivisions of Preventive Medicine and Aging, Brigham and Women’s Hospital, Harvard Medical School, Boston, Mass; ei3 Drug Safety, Waltham, Mass. ABSTRACT PURPOSE: We estimated the absolute risks of treatment termination and incidence of adverse liver outcomes among all commonly used oral antifungal treatments for superficial dermatophytosis and onychomycosis. METHODS: MEDLINE, EMBASE, and Cochrane Library were searched to identify randomized and nonrandomized controlled trials, case series, and cohort studies published before December 31, 2005. Two reviewers independently applied selection criteria, performed quality assessment, and extracted data. Treatment arms with the same regimen in terms of drug, type (continuous or intermittent), and dosage were combined to estimate the risk of an outcome of interest. RESULTS: We identified 122 studies with approximately 20,000 enrolled patients for planned comparison. The pooled risks (95% confidence intervals) of treatment discontinuation resulting from adverse reactions for continuous therapy were 3.44% (95% confidence interval [CI], 2.28%-4.61%) for terbinafine 250 mg/day; 1.96% (95% CI, 0.35%-3.57%) for itraconazole 100 mg/day; 4.21% (95% CI, 2.33%-6.09%) for itraconazole 200 mg/day; and 1.51% (95% CI, 0%-4.01%) for fluconazole 50 mg/day. For intermittent therapy, the pooled risks were as follows: pulse terbinafine: 2.09% (95% CI, 0%-4.42%); pulse itraconazole: 2.58% (95% CI, 1.15%-4.01%); intermittent fluconazole 150 mg/week: 1.98% (95% CI, 0.05%-3.92%); and intermittent fluconazole 300 to 450 mg/week: 5.76% (95% CI, 2.42%-9.10%). The risk of liver injury requiring termination of treatment ranged from 0.11% (continuous itraconazole 100 mg/day) to 1.22% (continuous fluconazole 50 mg/day). The risk of having asymptomatic elevation of serum transaminase but not requiring treatment discontinuation was less than 2.0% for all treatment regimens evaluated. CONCLUSION: Oral antifungal therapy against superficial dermatophytosis and onychomycosis, including intermittent and continuous terbinafine, itraconazole, and fluconazole, was associated with a low incidence of adverse events in an immunocompetent population. © 2007 Elsevier Inc. All rights reserved. KEYWORDS: Adverse reactions; Antifungal agents; Dermatophytosis; Fluconazole; Harms; Itraconazole; Metaanalysis; Onychomycosis; Terbinafine; Safety; Systematic review

Superficial dermatophytosis and onychomycosis are the 2 most common dermatologic diseases in the United States. It was estimated that up to 20% of the population had derThis study was supported in part by National Institutes of Health grant RO-1 DK62322 and the Harvard Pharmacoepidemiology Program. Requests for reprints should be addressed to Arnold K. Chan, MD, ScD, Department of Epidemiology, Harvard School of Public Health, 677 Huntington Ave, Boston MA 02115. E-mail address: [email protected].

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2007.03.021

matophytic infections and that the prevalence of onychomycosis was 8.7%.1,2 A recent 12-center North American study found a prevalence of confirmed onychomycosis of 13.8%.3 Meanwhile, analyses of office-based physician visits for fungal skin infections recorded in the 1990 to 1994 National Ambulatory Medical Care Survey revealed that during this period there were 4.1 million visits for cutaneous fungal diseases, of which 60% were visits to internists or family physicians.4

792

The American Journal of Medicine, Vol 120, No 9, September 2007

Oral antifungal therapy with newer agents, such as terSelection Criteria for Studies binafine, itraconazole, and fluconazole, is the treatment of We included randomized and nonrandomized controlled choice for onychomycosis and dermatophytosis not retrials, case series, and cohort studies that enrolled patients sponding to topical therapy. Their safety profiles deserve having superficial dermatophytosis (tinea pedis, tinea mascrutiny because a highly favorable risk– benefit ratio is nus, tinea corpora, and tinea cruris) or onychomycosis, aged expected for treating patients with 18 years or more, and receiving common and non–life-threatening oral antifungal therapy for 2 or dermatologic conditions. Adverse more weeks. Only English articles CLINICAL SIGNIFICANCE reactions associated with the use published before December 31, of oral antifungal agents are usu2005, were included. Pharmacoki● Oral terbinafine, itraconazole, and flually mild, transient, and reversible netic studies, studies focusing on conazole were associated with a low inafter discontinuation of therapy. drug– drug interactions, and studcidence of adverse events in patients The most frequently reported adies that enrolled patients with huwith skin and nail fungal infection. verse effects are associated with man immunodeficiency virus ingastrointestinal system, followed ● Pulse therapy was associated with a fection/acquired immunodeficiency by skin complaints and headlower risk of discontinuation because of syndrome, cancer, hematologic aches.5 Although reported cases of malignancy; those who underwent adverse reactions than continuous symptomatic hepatic injury are bone marrow or organ transplantherapy. rare, there have been case reports tation; and those who received im● Fluconazole seemed to have a lower risk of serious liver injuries leading to munosuppressive therapy were of treatment discontinuation because of liver failure, transplantation, or excluded. We also excluded studdeath.6 Although individual readverse events. ies enrolling patients with tinea views examining the safety proversicolor (treatment duration ● The incidence of adverse events infiles of continuous oral antifungal mostly ⬃1 week) and tinea capitis creased as the dosage of antifungal therapy have been reported,7–10 (prevalent in children). We exagents increased. there has been no systematic evalcluded follow-up studies in which uation of safety profiles among patients who completed a full various regimens, including new course of oral antifungal treatment intermittent therapy, which was were evaluated for an additional period of time to assess assumed to have additional advantages of improved patient relapse rate. compliance, lower cost, and reduced risk of systematic Studies including oral continuous or intermittent therapy adverse reactions.11 We conducted a meta-analysis to comof terbinafine, itraconazole, or fluconazole, with or without bine safety information from all published studies of antitopical agents, as one of the treatment arms were included, fungal therapy for superficial dermatophytosis and onychoregardless of antifungal dosage, length of therapy, or conmycosis to obtain absolute risk estimates associated with tinuous or pulse/intermittent therapy. Treatment arms inthese treatment regimens. volving sequential oral antifungal therapy were included only if clear safety end points were reported for each treatment phase of the trials. METHODS We followed the recommendations of the Quality of Reporting of Meta-analyses conference in conducting this systematic review.12

Literature Search Strategy We searched MEDLINE, EMBASE, and Cochrane Library (which includes the Cochrane Database of Systematic Reviews; the Database of Abstracts of Reviews of Effects; and the Cochrane Register of Controlled Trials) for relevant articles. Medical Subjects Heading (MeSH) terms used for keyword and text word searching included antifungals, terbinafine, itraconazole, fluconazole, onychomycosis, and dermatophytosis. The references of 8 systematic review articles on treatment for dermatophytosis and onychomycosis were examined to identify additional studies that were not found in the computerized databases.13–20 Additional reports were identified from reference lists of the identified articles.

Safety Outcomes The primary outcome of interest in this meta-analysis was the cumulative incidence of patients who withdrew from the study because of adverse reactions. The secondary outcomes of interest were the cumulative incidence of patients stopping treatment because of elevation of serum transaminase levels and the cumulative incidence of patients developing elevation of serum transaminase levels during treatment but not requiring discontinuation. We did not define specific cutoff values for serum transaminase levels that warranted treatment termination because different criteria were used in different studies.

Data Extraction Two physician reviewers independently evaluated each identified study and abstracted relevant characteristics, including the quality of the studies. Disagreement on specific

Chang et al

Safety of Oral Antifungals

studies between the 2 reviewers was resolved through discussion. Abstracted information included study characteristics (author, year the study was published, study design, treatment regimen, dose, and duration), patient characteristics (dermatophytosis or onychomycosis, percentage of males, and mean age), sample size, efficacy outcomes, proportions of patients who withdrew because of adverse events, and more specifically because of elevation of serum transaminase levels, as well as the proportion of patients developing elevation of serum transaminase levels but not requiring treatment discontinuation. For each treatment arm the number of patients who received at least 1 dose of oral antifungal agents was used as the denominator of the cumulative incidence. Studies that did not report this safety outcome information were excluded from analysis. We reported the summary results for regimens with data from 3 or more studies.

Statistical Analysis We anticipated large variations in trial design in terms of dosage, treatment duration, continuous or pulse/intermittent regimen, with or without concomitant topical therapy. Instead of conducting a meta-analysis using relative effect measures, such as risk difference, relative risk, or odds ratio that required the same contrast of comparison in all studies, we used absolute risk as our outcome of interest because it was not restricted by the comparative arms and its interpretation has direct clinical meaning. We used the beta-binomial model to capture the variation across studies and the maximum likelihood method to estimate the mean pooled event risks.21 From all the eligible studies, we combined all treatment arms with the same regimen in terms of type (continuous, intermittent) and dosage and obtained the summary estimates of cumulative incidence and their 95% confidence intervals (CIs). In the situation in which few adverse events occurred and the beta-binomial distribution collapsed to a simple binomial distribution, we calculated the Wald CIs.22 We used the adjusted Wald method to calculate the point estimates and 95% CIs for those risk estimates corresponding to no event.23 For rare outcomes, lower bounds of the CIs were set to be no smaller than zero. The main analytic results were obtained from combining data from randomized controlled trials only. In the auxiliary analysis, we added data from nonrandomized controlled trials, case series, and cohort studies. We also stratified the analysis by disease indication (dermatophytosis or onychomycosis).

RESULTS We identified 299 studies that reported oral antifungal treatment for dermatophytosis or onychomycosis from computerized literature databases and reference lists of systematic reviews and identified articles (Figure 1); 214 of them were reported in English and were retrievable for review. Two physicians independently reviewed all reports. A total of 77 studies were excluded for 1 of the following reasons: age less than 18 years for study subjects; therapies were given for

793 nonsuperficial dermatophytosis or onychomycosis (tinea imbricata, Pityrosporum folliculitis, cutaneous candidiasis, and systematic mycosis); studies of ketoconazole or griseofulvin only; immunocompromised patients or patients with cancer; pharmacokinetic or mycologic studies; same study cohorts or reports for trial follow-up; and treatment duration of less than 2 weeks. Among the remaining studies, relevant safety information was reported in 122 studies, and they were included in the meta-analysis (Appendix). These 122 studies, published from 1987 to 2005 with the number of study subjects ranging from 5 to 1608, included a total of 19,298 patients. Seventyseven of the studies were randomized controlled trials. Most of these trials were head-to-head comparisons of a variety of treatment regimens, and 15 studies included a placebo arm. The characteristics of the randomized controlled trial arms that we evaluated are summarized in Table 1. Most commonly used continuous regimens in the published randomized controlled trials were terbinafine 250 mg/day (41 arms, 3135 patients); itraconazole 100 mg/day (19 arms, 1002 patients) and 200 mg/day (12 arms, 2145 patients); and fluconazole 50 mg/day (3 arms, 235 patients). Terbinafine 500 mg/day (5 arms, 359 patients) or itraconazole 400 mg/day (15 arms, 766 patients) for 1 week and then repeated every month, and intermittent fluconazole 150 mg (7 arms, 514 patients) or 300 to 450 mg (3 arms, 468 patients) once weekly were the most commonly reported intermittent therapy regimens. The mean age of patients enrolled in the trials ranged from 27 to 72 years, and the proportion of male patients ranged from 14% to 100%. There was substantial variability in the reported risk of safety outcomes among study arms receiving the same antifungal regimens. The summarized treatment termination and safety information for all commonly reported regimens of continuous and intermittent oral antifungal therapy are presented in Tables 2 and 3. For continuous oral antifungal therapy, the pooled risks (95% CIs) of treatment discontinuation because of adverse reactions were 3.44% (95% CI, 2.28%-4.61%) for terbinafine 250 mg/day; 1.96% (95% CI, 0.35%-3.57%) for itraconazole 100 mg/day; 4.21% (95% CI, 2.33%6.09%) for itraconazole 200 mg/day; and 1.51% (95% CI, 0%-4.01%) for fluconazole 50 mg/day. For intermittent or pulse therapy, the pooled risks were as follows: pulse terbinafine: 2.09% (95% CI, 0%-4.42%); pulse itraconazole: 2.58% (95% CI, 1.15%-4.01%); intermittent fluconazole 150 mg/week: 1.98% (95% CI, 0.05%-3.92%); and intermittent fluconazole 300 to 450 mg/week: 5.76% (95% CI, 2.42%-9.10%). The incidence of liver injury associated with oral antifungal therapy was less than 2% in general. For the risks of having elevated serum transaminase levels that required treatment termination, the pooled risk estimates for continuous therapy ranged from 0.11% (itraconazole 100 mg/day) to 1.22% (fluconazole 50 mg/day). The pooled risk estimates for pulse therapy ranged from 0.39% (fluconazole 150 mg/week and itraconazole 400 mg/day) to 0.85% (fluconazole 300-450 mg/ week). The pooled risks of developing elevated serum

794

The American Journal of Medicine, Vol 120, No 9, September 2007

Figure 1

Literature search and selection of published reports.

transaminase levels not requiring treatment discontinuation was on the order of 1.5% for continuous regimens and 1% for intermittent regimens evaluated. The pooled probabilities of the 3 safety end points for the placebo arms in double-blind randomized controlled trials (15 arms, 891 patients) were 3.25% (95% CI, 1.48%5.02%), 1.01% (95% CI, 0%-2.19%), and 1.15% (95% CI, 0%-3.46%), respectively. These estimates might be interpreted as inflated estimates for the “background rates” of these events among patients with superficial dermatophytosis or onychomycosis because of the intensive laboratory monitoring in randomized controlled trials. None of the above treatment regimens had an incidence of adverse events substantially higher than these background rates. We stratified the risk estimates according to indication for oral antifungal therapy. Continuous terbinafine 250 mg/ day and intermittent fluconazole 150 mg/week were used in treating both dermatophytosis and onychomycosis. The pooled estimates of discontinuing continuous terbinafine because of adverse reactions were 2.72% (95% CI, 1.02%-

4.42%) for 2 to 6 weeks against dermatophytosis (11 arms, 623 patients) and 3.74% (95% CI, 2.28%-5.21%) for 8 to 48 weeks against onychomycosis (30 arms, 2512 patients). The probabilities of developing elevations of serum transaminase levels that required treatment stopping were 0.31% (95% CI, 0%-0.74%) for 2 to 6 weeks and 0.44% (95% CI, 0.13%-0.76%) for 8 to 48 weeks. For intermittent fluconazole 150 mg/week, the pooled risks of treatment termination attributable to all safety concern and to adverse liver outcomes were 0.31% (95% CI, 0%-0.93%) and 0.59% (95% CI, 0%-1.43%) for 2 to 6 weeks (3 arms, 320 patients), and 4.12% (95% CI, 1.33%-6.92%) and 1.03% (95% CI, 0%-2.45%) for 8 weeks or longer (4 arms, 194 patients). We identified 10 nonrandomized controlled clinical trials and 35 case series or cohort study reports with sufficient safety information reported. In the auxiliary analysis that added data from these studies, the safety profiles of different oral antifungal regimens generally followed the same order as results from the main analysis of randomized controlled trials (Tables 2 and 3).

Chang et al

Characteristics of the Included Randomized Controlled Trial Arms with Commonly Used Oral Antifungal Regimens Range of Patient Demographics

Drug Continuous T T I I I F Intermittent T I F F F Placebo

Range of Reported Safety Outcomes

Treatment Termination Because of Adverse Reaction (%)

Treatment Termination Because of Elevated Serum Transaminase (%)

Transaminase Elevation Did Not Require Treatment Termination (%)

0

0%–2.56% 0%–5.26%

0%–10% 0%–2.15% 0%–4.44% 0%–5.56% 0%–5.91% 0%–4.30%

Dosage

Indication

No. of RCT Arms

No.

Mean Age (y)

Proportion of Males

250 mg/d 250 mg/d 100 mg/d 100 mg/d 200 mg/d 50 mg/d

D O D O O D

11 30 15 4 12 3

18–184 5–186 10–303 10–45 27–1212 19–123

29–47 36–68 27–46 38–48 39–49 39–43

0.58–1.00 0.46–0.97 0.43–0.88 0.26–0.57 0.30–0.74 0.70–0.72

0%–10.00% 0%–12.05% 0%–6.06% 0%–13.89% 0%–7.69% 0%–5.26%

500 mg/d ⫻ 1⁄4wk* 400 mg/d ⫻ 1⁄4wk* 150 mg/wk† 150 mg/wk† 300–450 mg/wk†

O

5

5–153

39–64

0.48–0.96

0%–5.23%

0%–1.05%

0

O

15

12–126

36–71

0.12–0.96

0%–6.35%

0%–3.70%

0%–6.12%

D O O D, O

3 4 3 15

79–122 16–89 88–288 22–104

42 43–49 48 40–72

0.64–0.82 0.32–0.73 0.73 0.58–1.00

0%–0.82% 0%–4.55% 3.47%–10.23% 0%–12.50%

0%–1.47% 0%–1.45% 0%

0 0%–2.27% 0.35%–2.27% 0%–12.50%

Safety of Oral Antifungals

Table 1

0%–0.82% 0%–4.55% NA 0%–13.64%

T ⫽ terbinafine; I: itraconazole; F ⫽ fluconazole; D ⫽ dermatophytosis; O ⫽ onychomycosis; RCT ⫽ randomized controlled trial; NA ⫽ not available. *Administered for 1 week and then repeated every 4 weeks. †Once weekly.

795

796 Table 2

The American Journal of Medicine, Vol 120, No 9, September 2007 Pooled Risk Estimates of Safety Outcomes for Various Regimens of Continuous Oral Antifungal Therapy

Drug

Regimen

Terbinafine

250 mg/d

Itraconazole

100 mg/d

Itraconazole

200 mg/d

Fluconazole

50 mg/d

Treatment Termination Because of Adverse Effects (%)

Transaminase Elevation Required Treatment Termination (%)

Transaminase Elevation Did Not Require Treatment Termination (%)

Study Design

No. of Arms/ Patients Included

Pooled Estimate

95% CI

Pooled Estimate

95% CI

Pooled Estimate

95% CI

RCT All RCT All RCT All RCT All

41/3135 57/3563 19/1002 29/1307 12/2145 14/2183 3/235 5/879

3.44 3.45 1.96 1.91 4.21 3.94 1.51 2.30

2.28–4.61 2.39–4.51 0.35–3.57 0.59–3.23 2.33–6.09 2.17–5.70 0.00–4.01 0.93–3.66

0.34 0.35 0.11 0.08 0.70 0.68 1.22 0.59

0.09–0.60 0.09–0.62 0.00–0.32 0.00–0.25 0.33–1.06 0.33–1.04 0.00–5.30 0.00–1.79

0.70 0.70 1.24 1.21 1.90 1.98 1.63 2.16

0.08–1.32 0.09–1.30 0.00–2.64 0.15–2.28 0.13–3.68 0.36–3.61 0.00–5.02 0.02–4.30

RCT ⫽ randomized controlled trial; CI ⫽ confidence interval; All ⫽ including randomized and nonrandomized controlled trials, case series, and cohort studies; NA ⫽ not available because not reported in the published literature. ⴱAdjusted Wald method for point estimate and CI.

DISCUSSION Several observational studies examined the safety profiles of different oral antifungal agents for treating dermatophytosis or onychomycosis.7,8,10,24 They concluded that the incidence of serious adverse reaction associated with the use of new oral antifungal agents was in a low range. We conducted this meta-analysis to estimate the average cumulative incidences of 3 important safety end points among commonly used oral antifungal regimens from published reports. With regard to both continuous and intermittent therapy, the risk of treatment discontinuation because of adverse reaction ranged from 1.5% (fluconazole 50 mg/day) to 4.2% (itraconazole 200 mg/day). The risk of liver injury

Table 3

requiring or not requiring treatment discontinuation was less than 2.0% for all evaluated treatment regimens. None of the above treatment regimens had risks of adverse events substantially higher than the background risks estimated from the placebo arms in double-blind randomized controlled trials. These findings, in accordance with the results from prior studies, suggested that oral antifungal therapy was well tolerated and safe in a general immunocompetent population. In this systematic review with a comprehensive literature search, we identified approximately 50 studies with a total of 5586 patients, including at least 1 treatment arm of intermittent antifungal therapy. Because of the multiplicity

Pooled Risk Estimates of Safety Outcomes for Various Regimens of Intermittent Oral Antifungal Therapy

Drug

Regimen

Terbinafine

500 mg/d ⫻ 1⁄4wk

Itraconazole

400 mg/d ⫻ 1⁄4wk

Fluconazole

150 mg/wk

Fluconazole

300–450 mg/wk

Study Design

No. of Arms/ Patients Included

Treatment Termination Because of Adverse Effects (%)

Transaminase Elevation Required Treatment Termination (%)

Transaminase Elevation Did Not Require Treatment Termination (%)

Pooled Estimate

95% CI

Pooled Estimate

95% CI

Pooled Estimate

95% CI

RCT

5/359

2.09

0.00–4.42

0.56

0.00–1.33

0.92*

0.00–2.20*

All RCT

6/514 15/766

2.97 2.58

0.69–5.24 1.15–4.01

0.58 0.39

0.00–1.24 0.00–0.93

1.08 1.04

0.00–3.43 0.00–2.27

All RCT All RCT All

25/2856 7/514 13/899 3/468 5/507

2.22 1.98 1.57 5.76 4.99

1.04–3.40 0.05–3.92 0.40–2.73 2.42–9.10 2.21–7.77

0.24 0.39 0.33 0.85 0.79

0.00–0.67 0.00–0.93 0.00–0.71 0.02–1.69 0.02–1.56

0.75 0.81 0.58 NA 0.38

0.07–1.44 0.00–2.05 0.00–1.47 NA 0.00–0.91

RCT ⫽ randomized controlled trial; CI ⫽ confidence interval; All ⫽ including randomized and nonrandomized controlled trials, case series, and cohort studies; NA ⫽ not available because not reported in the published literature. *Adjusted Wald method for point estimate and CI.

Chang et al

Safety of Oral Antifungals

of the treatment regimens in terms of drug, dosage, and type of therapy, we could not perform 1-to-1 direct comparisons between each of these individual regimens. We found that continuous therapy, particularly itraconazole 200 mg/day, seemed to be associated with a higher risk of discontinuation because of adverse reactions and liver injury. Compared with continuous itraconazole 200 mg/day, pulse itraconazole seemed to be associated with a lower risk of discontinuation and liver toxicity. Furthermore, continuous terbinafine 250 mg/day had a better hepatic safety profile compared with high-dose continuous itraconazole. However, we did not find a reduction of the risk of transaminase elevation for pulse terbinafine therapy. From 10 randomized controlled trials that included fluconazole as one of the treatment arms, we found that fluconazole, in continuous or intermittent use, might have a lower risk of treatment discontinuation because of adverse events compared with other antifungal agents. As for the dose effects, we found that the incidence of adverse events increased markedly as the dosage of continuous itraconazole increased from 100 mg/day to 200 mg/ day and the dosage of intermittent fluconazole increased from 150 mg/day to 300 to 450 mg/day. This finding was consistent with clinical observations that risk became higher during high-dose antifungal treatment. Furthermore, as we stratified our analysis on the indication and compared shorter duration of use of oral antifungal agents for dermatophytosis with longer duration of use for onychomycosis, we did not find strong evidence that longer duration of use was associated with an increased risk of adverse liver outcomes. This was supported by the fact that adverse reactions to oral antifungal agents were idiosyncratic in nature, which tended to occur within the first few weeks of treatment, and their occurrence is nonlinear with time.25 Our main analytic results came from data synthesis of randomized controlled trials, which were supposed to have high quality for reporting safety outcomes. However, randomized controlled trials were different from routine clinical practice because of their highly selected patient population and intensive laboratory monitoring. To obtain estimates that may reflect real-life practice, we attempt to capture more of the heterogeneity of clinical practice through the auxiliary analysis that included case series and cohort studies. Nevertheless, we did not find a substantial change in the results, although most of the pooled estimates decreased and the CIs became narrower, probably because of a less rigorous laboratory monitoring plan adopted in these studies. Liver function of enrolled patients was monitored every 2 weeks for oral antifungal use of 2 to 6 weeks and at monthly intervals for use of 8 weeks or longer in the clinical trials. Because of limited information about the time of occurrence of adverse events in the published reports,26 no conclusive recommendation about the optimal liver function test monitoring schedule could be derived for patients receiving oral antifungal agents. In several articles, abnormal

797 transaminase levels were mostly detected during the interval of week 2 to week 6 after initiation of therapy. In addition, several limitations of our study should be considered. First, we might not have included all published articles because we only searched the MEDLINE, EMBASE, and Cochrane Library databases, and the reference lists of the identified articles and systematic reviews, and limited study enrollment to the English literature. However, we consider it unlikely that the association between treatment with antifungal regimens and adverse outcome differs in non-included studies. Second, monitoring of patients and the quality in reporting the safety outcomes varied among studies with different study designs. Although different trials might use different cutoff points of liver enzyme levels for treatment discontinuation, such discrepancy reflects clinical practice. Third, instead of conducting a formal test of heterogeneity, we explored the variation of the reported outcomes by a priori analysis plan stratifying on treatment regimen, dosage, and indication. However, there was not enough remaining variation in age and gender distribution after stratifying on regimen and indication; thus, we could not further explore whether these 2 factors contribute to the heterogeneity of the reported safety outcomes.

CONCLUSION In this meta-analysis of 122 studies with approximately 20,000 enrolled patients, we found that oral antifungal therapy against superficial dermatophytosis and onychomycosis, including intermittent and continuous terbinafine, itraconazole, and fluconazole, was associated with a low risk of adverse events in an immunocompetent population.

ACKNOWLEDGMENT The authors thank Dr Chia-Yi Lu for review of the antifungal studies.

References 1. Vander Straten MR, Hossain MA, Ghannoum MA. Cutaneous infectious dermatophytosis, onychomycosis, and tinea versicolor. Infect Dis Clin N Am. 2001;17:87-112. 2. Elewsk BE, Charif MA. Prevalence of onychomycosis in patients attending a dermatology clinic in northeastern Ohio for other conditions. Arch Dermatol. 1997;133:1172-1173. 3. Ghannoum MA, Hajjeh RA, Scher R, et al. A large-scale North American study of fungal isolates from nails: the frequency of onychomycosis, fungal distribution, and antifungal susceptibility patterns. J Am Acad Dermatol. 2000;43:641-648. 4. Smith ES, Fleischer AB, Feldman ST, Williford PM. Characteristics of office-based physician visits for cutaneous fungal infections: an analysis of 1990 to 1994 National Ambulatory Medical Care Survey Data. Cutis. 2002;69:191-202. 5. Gupta AK, Ryder JE. The use of oral antifungal agents to treat onychomycosis. Dermatol Clin. 2003;21:469-479. 6. Ajit C, Suvannasankha A, Zaeri N, Munoz SJ. Terbinafine-associated hepatotoxicity. Am J Med Sci. 2003;325:292-295. 7. O’Sullivan DP, Needham CA, Bangs A, et al. Postmarketing surveillance of oral terbinafine in the UK: report of a large cohort study. Br J Clin Pharmacol. 1996;42:559-565.

798

The American Journal of Medicine, Vol 120, No 9, September 2007

8. Hall M, Monka C, Krupp P, O’Sullivan D. Safety of oral terbinafine: results of a postmarketing surveillance study in 25,884 patients. Arch Dermatol. 1997;133:1213-1219. 9. Gupta AK, Chwetzoff E, Del Rosso F, Baran R. Hepatic safety of itraconazole. J Cutan Med Surg. 2002;210-213. 10. Garcia-Rodriguez LA, Castellsague AD, Perez-Gutthann S, Stricker BHC. A cohort study on the risk of acute liver injury among users of ketoconazole and other antifungal drugs. Br J Clin Pharmacol. 1999; 48:847-852. 11. Alpsoy E, Yilmaz E, Basaran E. Intermittent therapy with terbinafine for dermatophyte toe-onychomycosis: a new approach. J Dermatol. 1996;23:259-262. 12. Moher D, Cook DJ, Eastwood S, et al. Improving the quality of reports of meta-analyses of randomised controlled trials: the QUOROM statement. Quality of Reporting of Meta-analyses. Lancet. 1999;354:1896-1900. 13. Epstein E. How often does oral treatment of toenail onychomycosis produce a disease-free nail? An analysis of published data. Arch Dermatol. 1998;134:1551-1554. 14. Bell-Syer SEM, Hart R, Crawford F, et al. A systematic review of oral treatments for fungal infectious of the skin of the feet. J Dermatol Treat. 2001;12:69-74. 15. Cribier BJ, Paul C. Long-term efficacy of antifungals in toenail onychomycosis: a critical review. Br J Dermatol. 2001;145:446-452. 16. Haugh M, Helou S, Boissel JP, Cribier BJ. Terbinafine in fungal infections of the nail: a meta-analysis of randomized clinical trials. Br J Dermatol. 2002;147:118-121. 17. Crawford F, Young P, Godfrey C, et al. Oral treatments for toenail onychomycosis. A systematic review. Arch Dermatol. 2002;138:811-816.

18. Bell-Syer SEM, Hart R, Crawford F, et al. Oral treatments for fungal infections of the skin of the foot. Cochrane Database Syst Rev. 2002; (2):CD003584. 19. Krob HA, Fleischer AB, D’Agostino R, Feldman SR. Terbinafine is more effective than itraconazole in treating toenail onychomycosis: results from a meta-analysis of randomized controlled trials. J Cutan Med Surg. 2003;306-311. 20. Gupta AK, Ryder JE, Johnson AM. Cumulative meta-analysis of systematic antifungal agents for the treatment of onychomycosis. Br J Dermatol. 2004;150:537-544. 21. Chuang-Stein C. An application of the beta-binomial model to combine and monitor medical event rates in clinical trials. Drug Inf J. 1993;27:515-523. 22. Daniel W. Probability distribution. In: Biostatistics: A Foundation for Analysis in the Health Sciences. 7th ed. New York: Wiley & Sons; 1999;97. 23. Agresti A, Coull B. Approximate is better than ‘exact’ for interval estimation of binomial proportions. Am Stat. 1998;52:119-126. 24. Castellsague J, Garcia-Rodriguez LA, Duque A, Perez S. Risk of serious skin disorders among users of oral antifungals: a populationbased study. BMC Dermatol. 2002;2:14. 25. Chitturi S, Farrell GC. Drug-induced liver disease. In: Schiff ER, Sorell MF, Maddrey WC, eds. Schiff’s Diseases of the Liver. 9th ed. Philadelphia: Lippincott, Williams & Wilkins; 2003:1063. 26. Ioannidis JPA, Evans SJW, Gotzsche PC, et al. Better reporting of harms in randomized trials: an extension of the CONSORT statement. Ann Intern Med. 2004;141:781-788.

Chang et al

Safety of Oral Antifungals

798.e1

APPENDIX Studies Included in Meta-analysis 1. Albanese G, Di Cintio R, Martini C, Nicoletti A. Short therapy for tinea unguium with terbinafine: four different courses of treatment. Mycoses. 1995:38:211-214. 2. Albreski DA, Gross EG. The safety of itraconazole in the diabetic population. J Am Podiatr Med Assoc. 1999:89:339-345. 3. Alpsoy E, Yilmaz E, Basaran E. Intermittent therapy with terbinafine for dermatophyte toe-onychomycosis: a new approach. J Dermatol. 1996:23:259-262. 4. Aman S, Akbar TM, Hussain I, et al. Itraconazole pulse therapy in the treatment of disto-lateral subungual onychomycosis. J Coll Physicians Surg Pak. 2003;13:618-620. 5. Arca, E, Tastan HB, Akar A, et al. An open, randomized, comparative study of oral fluconazole, itraconazole and terbinafine therapy in onychomycosis. J Dermatolog Treat. 2002:13:3-9. 6. Arenas R, Dominguez-Cherit J, Fernandez LM. Open randomized comparison of itraconazole versus terbinafine in onychomycosis. Int J Dermatol. 1995:34:138-143. 7. Arenas R, Fernandez G, Dominguez L. Onychomycosis treated with itraconazole or griseofulvin alone with and without a topical antimycotic or keratolytic agent. Int J Dermatol. 1991;30:586-589. 8. Assaf RR, Elewski BE. Intermittent fluconazole dosing in patients with onychomycosis: results of a pilot study. J Am Acad Dermatol. 1996;35(2 Pt 1):216-219. 9. Avner S, Nir N, Henri T. Combination of oral terbinafine and topical ciclopirox compared to oral terbinafine for the treatment of onychomycosis. J Dermatolog Treat. 2005;16:327-330. 10. Bahadir S, Inaloz HS, Alpay K, et al. Continuous terbinafine or pulse itraconazole: a comparative study on onychomycosis. J Eur Acad Dermatol Venereol. 2000;14:422-423. 11. Baldari U, Righini MG, Raccagni AA, et al. Comparative double blind, double dummy study on the efficacy and safety of fluconazole 100 mg/day versus terbinafine 250 mg/day in the treatment of dermatomycoses. G Ital Dermatol Venereol. 2000;135:229-235. 12. Baran R, Belaich S, Beylot C, et al. Comparative multicentre doubleblind study of terbinafine (250 mg per day) versus griseofulvin (1 g per day) in the treatment of dermatophyte onychomycosis. J Dermatolog Treat. 1997;8:93-97. 13. Baran R, Feuilhade M, Combernale P, et al. A randomized trial of amorolfine 5% solution nail lacquer combined with oral terbinafine compared with terbinafine alone in the treatment of dermatophytic toenail onychomycoses affecting the matrix region. Br J Dermatol. 2000;142:1177-1183. 14. Baudraz-Rosselet F, Rakosi T, Wili PB, Kenzelmann R. Treatment of onychomycosis with terbinafine. Br J Dermatol. 1992;126(Suppl 39):40-46. 15. Bonifaz A, Carrasco-Gerard E, Saul A. Itraconazole in onychomycosis: intermittent dose schedule. Int J Dermatol. 1997;36:70-72. 16. Boonk W, de Geer D, de Kreek E, et al. Itraconazole in the treatment of tinea corporis and tinea cruris: comparison of two treatment schedules. Mycoses. 1998;41:509-514. 17. Bourlond A, Lachapelle JM, Aussems J, et al. Double-blind comparison of itraconazole with griseofulvin in the treatment of tinea corporis and tinea cruris. Int J Dermatol. 1989;28:410-412. 18. Brautigam M, Nolting S, Schopf RE, Weidinger G. Randomised double blind comparison of terbinafine and itraconazole for treatment of toenail tinea infection. Seventh Lamisil German Onychomycosis Study Group. BMJ. 1995;311:919-922. 19. Chen J, Liao W, Wen H, et al. A comparison among four regimens of itraconazole treatment in onychomycosis. Mycoses. 1999;42: 93-96. 20. Chen X, Hiruma M, Shiraki Y, Ogawa H. Combination therapy of once-weekly fluconazole (100, 150, or 300 mg) with topical appli-

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

cation of ketoconazole cream in the treatment of onychomycosis. Jpn J Infect Dis. 2004;57:260-263. Cole GW, Stricklin G. A comparison of a new oral antifungal, terbinafine, with griseofulvin as therapy for tinea corporis. Arch Dermatol. 1989;125:1537-1539. De Backer M, De Vroey C, Lesaffre E, et al. Twelve weeks of continuous oral therapy for toenail onychomycosis caused by dermatophytes: a double-blind comparative trial of terbinafine 250 mg/ day versus itraconazole 200 mg/day. J Am Acad Dermatol. 1998;38 (5 Pt 3):S57-S63. De Bersaques J, Bjerke JR, Borelli S, et al. Comparison of oral fluconazole and topical clotrimazole in the treatment of fungal infections of the skin: European and American experience. Int J Dermatol. 1992;31(Suppl 2):21-26. De Cuyper C, Amblard P, Austad J, et al. Noncomparative study of fluconazole in the treatment of patients with common fungal infections of the skin. Int J Dermatol. 1992;31(Suppl 2):17-20. De Doncker P, Van Lint J, Dockx P, Roseeuw D. Pulse therapy with one-week itraconazole monthly for three or four months in the treatment of onychomycosis. Cutis. 1995;56:180-183. De Doncker PR, Scher RK, Baran RL, et al. Itraconazole therapy is effective for pedal onychomycosis caused by some nondermatophyte molds and in mixed infection with dermatophytes and molds: a multicenter study with 36 patients. J Am Acad Dermatol. 1997; 36(2 Pt 1):173-177. De Keyser P, De Backer M, Massart DL, Westelinck KJ. Twoweek oral treatment of tinea pedis, comparing terbinafine (250 mg/day) with itraconazole (100 mg/day): a double-blind, multicentre study. Br J Dermatol. 1994;130(Suppl 43):22-25. Degreef H, del Palacio A, Mygind S, et al. Randomized double-blind comparison of short-term itraconazole and terbinafine therapy for toenail onychomycosis. Acta Derm Venereol. 1999;79:221-223. Degreef H, Marien K, De Veylder H, et al. Itraconazole in the treatment of dermatophytoses: a comparison of two daily dosages. Rev Infect Dis. 1987;9(Suppl 1):S104-S108. del Palacio Hernandez A, Lopez Gomez S, Gonzalez Lastra F, et al. A comparative double-blind study of terbinafine (Lamisil) and griseofulvin in tinea corporis and tinea cruris. Clin Exp Dermatol. 1990;15:210-216. del Palacio A, Van Cutsem J, Lopez S, et al. Double-blind randomized comparative study on Tinea manuum and Tinea pedis treated with itraconazole and griseofulvin. Actas Dermo-Sifiliogr. 1993;84: 1-2, 45-53. Difonzo EM, Papini M, Cilli P, et al. A double-blind comparison of itraconazole and fluconazole in tinea pedis and tinea manuum. J Eur Acad Dermatol Venereol. 1995;4:148-152. Dobric I, Pasic A, Basta-Juzbasic A. Treatment of onychomycosis with oral terbinafine: an open clinical trial. Acta Dermatovenerol Croat. 1994;2:73-76. Drake L, Babel D, Stewart DM, et al. Once-weekly fluconazole (150, 300, or 450 mg) in the treatment of distal subungual onychomycosis of the fingernail. J Am Acad Dermatol. 1998;38(6 Pt 2):S87-S94. Drake LA, Shear NH, Arlette JP, et al. Oral terbinafine in the treatment of toenail onychomycosis: North American multicenter trial. J Am Acad Dermatol. 1997;37(5 Pt 1):740-745. Elewski BE, Scher RK, Aly R, et al. Double-blind, randomized comparison of itraconazole capsules vs. placebo in the treatment of toenail onychomycosis. Cutis. 1997;59:217-220. Evans EG, Sigurgeirsson B. Double blind, randomised study of continuous terbinafine compared with intermittent itraconazole in treatment of toenail onychomycosis. The LION Study Group. BMJ. 1999;318:1031-1035. Faergemann J, Anderson C, Hersle K, et al. Double-blind, parallelgroup comparison of terbinafine and griseofulvin in the treatment of toenail onychomycosis. J Am Acad Dermatol. 1995;32(5 Pt 1):750753. Faergemann J, Mork NJ, Haglund A, Odegard T. A multicentre (double-blind) comparative study to assess the safety and efficacy of

798.e2

40.

41.

42.

43.

44.

45.

46.

47.

48. 49.

50.

51.

52.

53.

54.

55.

56.

57. 58.

59.

The American Journal of Medicine, Vol 120, No 9, September 2007

fluconazole and griseofulvin in the treatment of tinea corporis and tinea cruris. Br J Dermatol. 1997;136:575-577. Farkas B, Paul C, Dobozy A, et al. Terbinafine (Lamisil) treatment of toenail onychomycosis in patients with insulin-dependent and noninsulin-dependent diabetes mellitus: a multicentre trial. Br J Dermatol. 2002;146:254-260. Finzi A, Cilli P. Italian multicentre trial comparing itraconazole with griseofulvin in the treatment of dermatomycoses. Preliminary results. J Eur Acad Dermatol Venereol. 1992;1(Suppl 1):S15-S18. Fischbein A, Haneke E, Lacner K, et al. Comparative evaluation of oral fluconazole and oral ketoconazole in the treatment of fungal infections of the skin. Int J Dermatol. 1992;31(Suppl 2):12-16. Fraki J, Heikkila HT, Kero MO, et al. An open-label, noncomparative, multicenter evaluation of fluconazole with or without urea nail pedicure for treatment of onychomycosis. Curr Ther Res Clin Exp. 1997;58:481-491. Galimberti R, Kowalczuk A, Flores V, Squiquera L. Onychomycosis treated with a short course of oral terbinafine. Int J Dermatol. 1996; 35:374-375. Ginter G, De Doncker P. An intermittent itraconazole 1-week dosing regimen for the treatment of toenail onychomycosis in dermatological practice. Mycoses. 1998;41:235-238. Goodfield MJ, Andrew L, Evans EG. Short term treatment of dermatophyte onychomycosis with terbinafine. BMJ. 1992;304:11511154. Goodfield MJ, Rowell NR, Forster RA, et al. Treatment of dermatophyte infection of the finger- and toe-nails with terbinafine (SF 86-327, Lamisil), an orally active fungicidal agent. Br J Dermatol. 1989;121:753-757. Goodfield MJ. Clinical results with terbinafine in onychomycosis. J Dermatolog Treat. 1990;1:(Suppl 2):55-57. Goodfield MJD, Evans EGV. Combined treatment with surgery and short duration oral antifungal therapy in patients with limited dermatophyte toenail infection. J Dermatolog Treat. 2000;11:259-262. Gupta AK, Gregurek-Novak T. Efficacy of itraconazole, terbinafine, fluconazole, griseofulvin and ketoconazole in the treatment of Scopulariopsis brevicaulis causing onychomycosis of the toes. Dermatology. 2001;202:235-238. Gupta AK, Konnikov N, Lynde CW. Single-blind, randomized, prospective study on terbinafine and itraconazole for treatment of dermatophyte toenail onychomycosis in the elderly. J Am Acad Dermatol. 2001;44:479-484. Gupta AK, Lynde CW, Konnikov N. Single-blind, randomized, prospective study of sequential itraconazole and terbinafine pulse compared with terbinafine pulse for the treatment of toenail onychomycosis. J Am Acad Dermatol. 2001;44:485-491. Gupta AK, Maddin S, Arlette J, et al. Itraconazole pulse therapy is effective in dermatophyte onychomycosis of the toenail: a doubleblind placebo-controlled study. J Dermatolog Treat. 2000;11:33-37. Haneke E, Abeck D, Ring R. Safety and efficacy of intermittent therapy with itraconazole in finger- and toenail onychomycosis: a multicentre trial. Mycoses. 1998;41:521-527. Haneke E, Tajerbashi M, De Doncker P, et al. Itraconazole in the treatment of onychomycosis: a double-blind comparison with miconazole. Dermatology. 1998;196:323-329. Haneke E, Tausch I, Brautigam M, et al. Short-duration treatment of fingernail dermatophytosis: a randomized, double-blind study with terbinafine and griseofulvin. LAGOS III Study Group. J Am Acad Dermatol. 1995;32:72-77. Hanifin JM, Tofte SJ. Itraconazole therapy for recalcitrant dermatophyte infections. J Am Acad Dermatol. 1988;18(5 Pt 1):1077-1080. Havu V, Brandt H, Heikkila H, et al. A double-blind, randomized study comparing itraconazole pulse therapy with continuous dosing for the treatment of toe-nail onychomycosis. Br J Dermatol. 1997; 136:230-234. Havu V, Heikkila H, Kuokkanen K, et al. A double-blind, randomized study to compare the efficacy and safety of terbinafine (Lamisil)

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71. 72.

73.

74.

75.

76.

77.

78.

79.

with fluconazole (Diflucan) in the treatment of onychomycosis. Br J Dermatol. 2000;142:97-102. Hay RJ, Clayton YM, Moore MK, Midgely G. An evaluation of itraconazole in the management of onychomycosis. Br J Dermatol. 1988;119:359-366. Hay RJ, Clayton YM, Moore MK, Midgely G. Itraconazole in the management of chronic dermatophytosis. J Am Acad Dermatol. 1990; 23(3 II Suppl):561-564. Hay RJ, Logan RA, Moore MK, et al. A comparative study of terbinafine versus griseofulvin in ‘dry-type’ dermatophyte infections. J Am Acad Dermatol. 1991;24(2 Pt 1):243-246. Hay RJ, McGregor JM, Wuite J, et al. A comparison of 2 weeks of terbinafine 250 mg/day with 4 weeks of itraconazole 100 mg/day in plantar-type tinea pedis. Br J Dermatol. 1995;132:604-608. Hofmann H, Brautigam M, Weidinger G, Zaun H. Treatment of toenail onychomycosis. A randomized, double-blind study with terbinafine and griseofulvin. LAGOS II Study Group. Arch Dermatol. 1995;131:919-922. Honeyman JF, Talarico FS, Arruda LHF, et al. Itraconazole versus terbinafine (LAMISIL(registered trademark)): which is better for the treatment of onychomycosis? J Eur Acad Dermatol Venereol. 1997; 9:215-221. Iozumi K, Hattori N, Adachi M, et al. Long-term follow-up study of onychomycosis: cure rate and dropout rate with oral antifungal treatments. J Dermatol. 2001;28:128-136. Jones HE, Zaias N. Double-blind, randomized comparison of itraconazole capsules and placebo in onychomycosis of toenail. Int J Dermatol. 1996;35:589-590. Khosravi AR, Mansouri P. Onychomycosis in Tehran, Iran: prevailing fungi and treatment with itraconazole.Mycopathologia. 2001;150: 9-13. Kim JA, Ahn KJ, Kim JM, Youn JI. Efficacy and tolerability of itraconazole in patients with fingernail onychomycosis: a 6-week pilot study. Curr Ther Res Clin Exp. 1995;56:1066-1075. Korting HC, Schafer-Korting M, Zienicke H, et al. Treatment of tinea unguium with medium and high doses of ultramicrosize griseofulvin compared with that with itraconazole. Antimicrob Agents Chemother. 1993;37:2064-2068. Kuokkanen K, Alava S. Fluconazole in the treatment of onychomycosis caused by dermatophytes. J Dermatolog Treat. 1992;3:115-117. Lecha M, Alsina M, Torres Rodriguez JM, et al. An open-label, multicenter study of the combination of amorolfine nail lacquer and oral itraconazole compared with oral itraconazole alone in the treatment of severe toenail onychomycosis. Curr Ther Res Clin Exp. 2002;63:366-379. Ling MR, Swinyer LJ, Jarratt MT, et al. Once-weekly fluconazole (450 mg) for 4, 6, or 9 months of treatment for distal subungual onychomycosis of the toenail. J Am Acad Dermatol. 1998;38(6 Pt 2):S95-S102. Mishra M, Panda P, Tripathy S, et al. An open randomized comparative study of oral itraconazole pulse and terbinafine pulse in the treatment of onychomycosis. Indian J Dermatol Venereol Leprol. 2005;71:262-266. Montero-Gei F, Perera A. Therapy with fluconazole for tinea corporis, tinea cruris, tinea pedis. Clin Infect Dis. 1992;(14 Suppl 1):S77S81. Montero-Gei F, Robles-Soto ME, Schlager H. Fluconazole in the treatment of severe onychomycosis. Int J Dermatol. 1996;35:587588. Nolting S, Brautigam M, Weidinger G. Terbinafine in onychomycosis with involvement by non-dermatophytic fungi. Br J Dermatol Suppl 1994;130:16-21. Noppakun N, Phuphaibool K. Treatment of dermatophytosis with new systemic antifungal agent, itraconazole. J Med Assoc Thai. 1992;75:99-103. Nozickova M, Koudelkova V, Kulikova Z, et al. A comparison of the efficacy of oral fluconazole, 150 mg/week versus 50 mg/day, in the

Chang et al

80. 81.

82.

83.

84.

85.

86.

87.

88.

89. 90.

91.

92. 93.

94.

95.

96. 97.

98.

99.

100.

Safety of Oral Antifungals

treatment of tinea corporis, tinea cruris, tinea pedis, cutaneous candidosis. Int J Dermatol. 1998;37:703-705. Nuijten ST, Schuller JL. Itraconazole in the treatment of tinea corporis: a pilot study. Rev Infect Dis. 1987;9(Suppl 1):S119-S120. Odom R, Daniel CR, Aly A. A double-blind, randomized comparison of itraconazole capsules and placebo in the treatment of onychomycosis of the toenail. J Am Acad Dermatol. 1996;35:110-111. Odom RB, Aly R, Scher RK, et al. A multicenter, placebo-controlled, double-blind study of intermittent therapy with itraconazole for the treatment of onychomycosis of the fingernail. J Am Acad Dermatol. 1997;36(2 Pt 1):231-235. Panagiotidou D, Kousidou T, Chaidemenos G, et al. A comparison of itraconazole and griseofulvin in the treatment of tinea corporis and tinea cruris: a double-blind study. J Int Med Res. 1992;20:392-400. Panconesi E, Difonzo E. Treatment of dermatophytoses and pityriasis versicolor with itraconazole. Rev Infect Dis. 1987;9(Suppl 1):S109S113. Parent D, Decroix J, Heenen M. Clinical experience with short schedules of itraconazole in the treatment of tinea corporis and/or tinea cruris. Dermatology. 1994;189:378-381. Pariser DM, Pariser RJ, Ruoff G, Ray TL. Double-blind comparison of itraconazole and placebo in the treatment of tinea corporis and tinea cruris. J Am Acad Dermatol. 1994;31(2 Pt 1):232-234. Pavlotsky F, Armoni G, Shemer A, Trau H. Pulsed versus continuous terbinafine dosing in the treatment of dermatophyte onychomycosis. J Dermatolog Treat. 2004;15:315-320. Piepponen T, Blomqvist K, Brandt H, et al. Efficacy and safety of itraconazole in the long-term treatment of onychomycosis. J Antimicrob Chemother. 1992;29:195-205. Rakosi T, Gerber M. Treatment of tinea with itraconazole: an open multicentre study. J Dermatolog Treat. 1995;6:35-38. Ramos-e-Silva M, Marques SA, Gontijo B, et al. Efficacy and safety of itraconazole pulse therapy: Brazilian multicentric study on toenail onychomycosis caused by dermatophytes. J Eur Acad Dermatol Venereol. 1998;11:109-116. Rigopoulos D, Katoulis AC, Ioannides D, et al. A randomized trial of amorolfine 5% solution nail lacquer in association with itraconazole pulse therapy compared with itraconazole alone in the treatment of Candida fingernail onychomycosis. Br J Dermatol. 2003;149:151156. Rongioletti F, Robert E, Tripodi S, Persi A. Treatment of onychomycosis with itraconazole. J Dermatolog Treat. 1992;2:145-146. Roseeuw D, Willemsen M, Kint RT, et al. Itraconazole in the treatment of superficial mycoses—a double-blind study vs. placebo. Clin Exp Dermatol.1990;15:101-104. Sanmano B, Hiruma M, Mizoguchi M, Ogawa H. Combination therapy consisting of week pulses of oral terbinafine plus topical application of terbinafine cream in the treatment of onychomycosis. J Dermatolog Treat. 2004;15:245-251. Saul A, Bonifaz A, Arias I. Itraconazole in the treatment of superficial mycoses: an open trial of 40 cases.Rev Infect Dis.1987;9(Suppl 1):S100-S103. Savin R. Successful treatment of chronic tinea pedis (moccasin type) with terbinafine (Lamisil). Clin Exp Dermatol. 1989;14:116-119. Savin RC, Zaias N. Treatment of chronic moccasin-type tinea pedis with terbinafine: a double-blind, placebo-controlled trial. J Am Acad Dermatol. 1990;23(4 Pt 2):804-807. Scher RK, Breneman D, Rich P, et al. Once-weekly fluconazole (150, 300, or 450 mg) in the treatment of distal subungual onychomycosis of the toenail. J Am Acad Dermatol. 1998;38(6 Pt 2):S77-S86. Schuller J, Remme JJ, Rampen FH, Van Neer FC. Itraconazole in the treatment of tinea pedis and tinea manuum: comparison of two treatment schedules. Mycoses. 1998;41:515-520. Segal R, Kritzman A, Cividalli L, et al. Treatment of Candida nail infection with terbinafine. J Am Acad Dermatol. 1996;35:958-961.

798.e3 101. Smith EB, Stein LF, Fivenson DP, Atillasoy ES. The safety of terbinafine in patients over the age of 60 years: a multicenter trial in onychomycosis of the feet. Int J Dermatol. 2000;39:861-864. 102. Smith SW, Sealy DP, Schneider E, Lackland D. An evaluation of the safety and efficacy of fluconazole in the treatment of onychomycosis. South Med J. 1995;88:1217-1220. 103. Stary A, Sarnow E. Fluconazole in the treatment of tinea corporis and tinea cruris. Dermatology. 1998;196:237-241. 104. Stengel F, Robles-Soto M, Galimberti R, Suchil P. Fluconazole versus ketoconazole in the treatment of dermatophytoses and cutaneous candidiasis. Int J Dermatol. 1994;33:726-729. 105. Svejgaard EL, Brandrup F, Kragballe K, et al. Oral terbinafine in toenail dermatophytosis. A double-blind, placebo-controlled multicenter study with 12 months’ follow-up. Acta Derm Venereol. 1997; 77:66-69. 106. Tang WYM, Chong LY, Leung CY, et al. Intermittent pulse therapy with itraconazole for onychomycosis. Experience in Hong Kong Chinese. Mycoses. 2000;43:35-39. 107. Tausch I, Brautigam M, Weidinger G, Jones TC. Evaluation of 6 weeks treatment of terbinafine in tinea unguium in a double-blind trial comparing 6 and 12 weeks therapy. The Lagos V Study Group. Br J Dermatol. 1997;136:737-742. 108. Tausch I, Decroix J, Gwiezdzinski Z, et al. Short-term itraconazole versus terbinafine in the treatment of tinea pedis or manus. Int J Dermatol. 1998;37:140-142. 109. Tosti A, Piraccini BM, Stinchi C, et al. Treatment of dermatophyte nail infections: an open randomized study comparing intermittent terbinafine therapy with continuous terbinafine treatment and intermittent itraconazole therapy. J Am Acad Dermatol. 1996;34:595-600. 110. Tuzun Y, Kotogyan A, Oguz O. Terbinafine: efficacy and safety in the treatment of dermatophytosis. Int J Dermatol. 1992;31:720-721. 111. van der Schroeff JG, Cirkel PK, Crijns MB, et al. A randomized treatment duration-finding study of terbinafine in onychomycosis. Br J Dermatol. 1992;126(Suppl 39):36-39. 112. van Hecke E, Van Cutsem J. Double-blind comparison of itraconazole with griseofulvin in the treatment of tinea pedis and tinea manuum. Mycoses. 1988;31:641-649. 113. Voravutinon V. Oral treatment of tinea corporis and tinea cruris with terbinafine and griseofulvin: a randomized double blind comparative study. J Med Assoc Thai. 1993;76:388-393. 114. Walsoe I, Stangerup M, Svejgaard E. Itraconazole in onychomycosis. Open and double-blind studies. Acta Derm Venereol. 1990;70:137140. 115. Wang DL, Wang AP, Li RY, Wang R. Therapeutic efficacy and safety of one-week intermittent therapy with itraconazole for onychomycosis in a Chinese patient population. Dermatology. 1999;199: 47-49. 116. Warshaw EM, Fett DD, Bloomfield HE, et al. Pulse versus continuous terbinafine for onychomycosis: a randomized, double-blind, controlled trial. J Am Acad Dermatol. 2005;53:578-584. 117. Watson A, Marley J, Ellis D, Williams T. Terbinafine in onychomycosis of the toenail: a novel treatment protocol. J Am Acad Dermatol. 1995;33(5 Pt 1):775-779. 118. White JE, Perkins PJ, Evans EG. Successful 2-week treatment with terbinafine (Lamisil) for moccasin tinea pedis and tinea manuum. Br J Dermatol. 1991;125:260-262. 119. Wishart JM. A double blind study of itraconazole vs griseofulvin in patients with tinea pedis and tinea manus. N Z Med J. 1994;107:126128. 120. Wong CK, Cho YL. Very short duration therapy with oral terbinafine for fingernail onychomycosis. Br J Dermatol. 1995;133:329-331. 121. Wu J, Wen H, Liao W. Small-dose itraconazole pulse therapy in the treatment of onychomycosis. Mycoses. 1997;40:397-400. 122. Zaias N, Serrano L. The successful treatment of finger Trichophyton rubrum onychomycosis with oral terbinafine. Clin Exp Dermatol. 1989;14:120-123.

The American Journal of Medicine (2007) 120, 799-806

CLINICAL RESEARCH STUDY

Impact of Cardiac Rehabilitation on Depression and Its Associated Mortality Richard V. Milani, MD, Carl J. Lavie, MD Department of Cardiology, Ochsner Medical Center, New Orleans, La. ABSTRACT PURPOSE: Depression following major cardiac events is associated with higher mortality, but little is known about whether this can be reduced through treatment including cardiac rehabilitation and exercise training. We evaluated the impact of cardiac rehabilitation on depression and its associated mortality in coronary patients. PATIENTS AND METHODS: We evaluated 522 consecutive coronary patients (381 men, 141 women; aged 64 ⫾ 10 years) enrolled in cardiac rehabilitation from January 2000 to July 2005 and a control group of 179 patients not completing rehabilitation. Depressive symptoms were assessed by questionnaire at baseline and following rehabilitation, and mortality was evaluated after a mean follow-up of 1296 ⫾ 551 days. RESULTS: Prevalence of depressive symptoms decreased 63% following rehabilitation, from 17% to 6% (P ⬍.0001). Depressed patients following rehabilitation had an over 4-fold higher mortality than nondepressed patients (22% vs 5%, P ⫽ .0004). Depressed patients who completed rehabilitation had a 73% lower mortality (8% vs 30%; P ⫽ .0005) compared with control depressed subjects who did not complete rehabilitation. Reductions in depressive symptoms and its associated mortality were related to improvements in fitness; however, similar reductions were noted in those with either modest or marked increases in exercise capacity. CONCLUSION: In patients following major coronary events, cardiac rehabilitation is associated with both reductions in depressive symptoms and the excess mortality associated with it. Moreover, only mild improvements in levels of fitness appear to be needed to produce these benefits on depressive symptoms and its associated mortality. © 2007 Elsevier Inc. All rights reserved. KEYWORDS: Coronary disease; Depression; Exercise; Mortality

Depression and depressive symptoms are prevalent in patients with coronary heart disease, comprising approximately 15%-20% of this population.1-4 Although its presence identifies patients at high risk for adverse cardiovascular outcomes including myocardial infarction (MI) and death, it remains underappreciated as a coronary risk factor.5-7 Identification and treatment of depression are especially important in patients following recent coronary events where its Presented in part to the Annual Scientific Sessions of the American Heart Association, Chicago, Illinois, November 2006. Requests for reprints should be addressed to Richard V. Milani, MD, Department of Cardiology, Ochsner Medical Center, 1514 Jefferson Highway, New Orleans, LA 70121. E-mail address: [email protected]

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2007.03.026

presence may be particularly pathogenic, conferring a greater than 4-fold increase in risk of death in the post-MI setting.8-12 Complicating this increased risk is the observation that major depression following MI runs a chronic course untreated, with more than 95% of patients remaining depressed at 6 months and up to 70% remaining depressed at 1 year.2,13 We and others have described the short-term benefits of cardiac rehabilitation and exercise training upon depressive symptoms in coronary heart disease patients following major cardiac events.14-19 Specifically, cardiac rehabilitation has been reported to reduce depressive symptoms and the prevalence of depression by 50%-70%. In a recent randomized trial, exercise training decreased depressive symptoms

800

The American Journal of Medicine, Vol 120, No 9, September 2007

as effectively as antidepressant medications.20,21 To date, tion consisting of 12 weeks of 36 educational and exercise however, the impact of reducing depressive symptoms utisessions. Exercise intensity was individually prescribed lizing cardiac rehabilitation has not been studied on major within 10 beats of the anaerobic threshold obtained on entry clinical events, including mortality. The present investigacardiopulmonary exercise testing. Each session consisted of tion studies the effects of rehabilitation upon depressive a 10-minute warm-up, 30 minutes of sustained aerobic exsymptoms and subsequent longercise, followed by a cool-down term mortality and evaluates the reperiod. Resistance training with lationship between exercise training light hand weights also was incorCLINICAL SIGNIFICANCE and improvements in fitness on this porated. At baseline, patients were behavior. instructed on the American Heart ● Following major coronary events, cardiac Association Step II diet with a rehabilitation and exercise training Mediterranean modification, and reduces the prevalence of depression, METHODS dieticians, exercise physiologists, improves levels of fitness, and signifinurses, and physicians frequently Patients cantly reduce all-cause mortality in encouraged patients to comply We retrospectively evaluated 522 coronary patients. with both the exercise and dietary consecutive patients who were reportions of the cardiac rehabilita● These data further support the importance ferred to, attended, and completed tion program. Daily lectures, as phase II cardiac rehabilitation of exercise training for both psychological well as group sessions about heart (treatment group) between Januhealth and overall survival. disease risk factors, general inforary 2000 and July 2005. All pamation about coronary disease tients entered the program 2-6 (symptoms, signs), and psychosoweeks after a coronary event, incial adaptations to this disease were given for patients and cluding acute MI (30%), coronary bypass surgery (35%), spouses. and percutaneous coronary intervention (44% of patients; Height, weight, body mass index, age, sex, fasting blood some patients had more than one clinical event). We further lipids, high sensitivity C-reactive protein (hs-CRP), and evaluated a control group of 179 patients who entered rehabilitation but dropped out within 2 weeks of entry (all peak oxygen consumption (peak VO2) were assessed at patients received ⬍5 sessions of rehabilitation) to evaluate baseline and again 1 week after completing rehabilitation as the impact of depressive symptoms in patients not underpreviously described.14,25,26 going rehabilitation. All patients completed questionnaires before rehabilitation, and the treatment group completed Statistical Analysis questionnaires after rehabilitation. None of these patients Statview software 5.0.1 (SAS Institute; Cary, NC) was used were being treated with antidepressive medications. Surfor statistical analysis. Results are mean ⫾ SD or frequenvival status was obtained January 1, 2006, after a mean cies expressed as percentages. Differences in continuous follow-up of 1296 ⫾ 551 days (range 109-2188 days) variables between 2 groups were assessed by paired Stufrom the National Death Index. The protocol was apdent’s t test or nonparametric tests as appropriate. Univarproved by the Institutional Review Committee at Ochsner iate relations between variables were assessed as partial Clinic Foundation. correlations. Two-tailed P ⱕ.05 was considered statistically significant. Logistic regression analysis was performed to Behavioral Testing determine independent predictors of mortality. Actuarial The Kellner Symptom Questionnaire is a 92-question assurvival analysis was used to compute cumulative hazard sessment validated to assess behavioral characteristics, inover time. cluding symptoms of depression, anxiety, somatization, and hostility, with a lower score being more favorable for each behavioral symptom (scores can range from 0-17 units).22 Depressive symptoms were considered to be significant when the depression score exceeded 6 (⬎7 for hostility and anxiety, ⬎8 for somatization) as previously described.15,17,19 The Medical Outcomes Short Form 36 survey was used to assess quality of life, with a high score indicating a more favorable quality-of-life trait.23

Protocol Protocol, data collection, and statistical analysis were performed as previously described.14,24 Patients were referred to and participated in outpatient phase II cardiac rehabilita-

RESULTS Baseline Characteristics The mean age of the 701 patients in treatment and control groups was 64 ⫾ 11 years, with 72% of the subjects male. The mean ejection fraction was 54 ⫾ 12%, and the mean peak VO2 was 16.2 ⫾ 5.2 mL/kg/min. Depressive symptoms were identified in 139 patients (20%) on entry into rehabilitation (26% controls, 17% treatment; P ⫽ .005). Table 1 highlights the baseline differences between treatment and control subjects. Table 2 outlines the baseline differences between treatment and control depressed patients.

Milani and Lavie

Exercise Reduces Depression-Related Mortality

Table 1 Baseline Differences in Treatment and Control Subjects on Entry into Cardiac Rehabilitation

Age (years) % Male Body mass index (kg/m2) Active smokers (%) Hypertensives (%) Diabetes (%) Married (%) Peak VO2 (Ml/kg/min) Total cholesterol (mg/dL) HDL ⫺ cholesterol (mg/dL) LDL ⫺ cholesterol (mg/dL) Triglycerides (mg/dL) hs-CRP (mg/dL) Ejection fraction (%) Depression Anxiety Hostility Somatization Quality of life

Treatment (n ⫽ 522)

Control (n ⫽ 179)

P-value

64 ⫾ 10 73% 28.5 ⫾ 5.1 1% 33% 22% 72% 16.6 ⫾ 5.2 167 ⫾ 38 41 ⫾ 13 97 ⫾ 36 148 ⫾ 84 5.7 ⫾ 9.3 54 ⫾ 12 3.3 ⫾ 4.0 4.1 ⫾ 4.6 2.5 ⫾ 3.5 6.6 ⫾ 3.9 102.9 ⫾ 17.8

63 ⫾ 12 69% 29.2 ⫾ 5.8 10% 78% 33% 75% 14.8 ⫾ 4.9 174 ⫾ 37 42 ⫾ 15 99 ⫾ 36 182 ⫾ 156 10.1 ⫾ 17.3 55 ⫾ 11 4.6 ⫾ 4.8 5.4 ⫾ 5.1 2.9 ⫾ 3.8 7.9 ⫾ 4.4 93.7 ⫾ 18.6

NS NS NS ⬍.0001 ⬍.0001 .01 NS ⬍.0001 .04 NS NS .0004 ⬍.0001 NS .001 .002 NS .0005 ⬍.0001

Treatment Group Characteristics and Effects of Cardiac Rehabilitation and Exercise Training

Table 3 Baseline Differences in Cardiac Rehabilitation Patients with and without Depression upon Study Entry

Age (years) % Male Body mass index (kg/m2) Active smokers (%) Hypertensives (%) Diabetes (%) Married (%) Peak VO2 (mL/kg/min) Total cholesterol (mg/dL) HDL ⫺ cholesterol (mg/dL) LDL ⫺ cholesterol (mg/dL) Triglycerides (mg/dL) hs-CRP (mg/dL) Ejection fraction (%) Depression Anxiety Hostility Somatization Quality of life

Depressed (n ⫽ 91)

Nondepressed (n ⫽ 431)

P-value

62 ⫾ 12 70% 28.6 ⫾ 5.4 4% 34% 24% 78% 16.0 ⫾ 5.2 169 ⫾ 34 40 ⫾ 12 96 ⫾ 30 162 ⫾ 76 6.5 ⫾ 10.1 53 ⫾ 13 10.7 ⫾ 3.4 10.7 ⫾ 4.9 7.0 ⫾ 4.9 10.2 ⫾ 3.9 83.5 ⫾ 15.0

65 ⫾ 10 74% 28.5 ⫾ 5.0 0.6% 31% 21% 71% 16.8 ⫾ 5.2 166 ⫾ 38 41 ⫾ 13 98 ⫾ 37 146 ⫾ 86 5.5 ⫾ 9.1 54 ⫾ 11 1.8 ⫾ 1.8 2.7 ⫾ 3.0 1.6 ⫾ 2.1 5.9 ⫾ 3.5 107.0 ⫾ 15.4

.017 NS NS .02 NS NS NS NS NS NS NS NS NS NS ⬍.0001 ⬍.0001 ⬍.0001 ⬍.0001 ⬍.0001

Depressive Symptoms and Survival

Of the 522 patients comprising the treatment group, depressive symptoms were identified in 91 (17%) patients on entry into rehabilitation. Table 3 highlights the baseline differences between treatment subjects based on the presence or absence of depressive symptoms. The effects of rehabilitation in the depressed cohort are described in Table 4. Following rehabilitation, the prevalence of depressive symptoms fell 63% (Figure 1) from 17% on entry, to 6% upon completion of the program (P ⬍.0001). Table 2 Baseline Differences in Depressed Treatment and Control Subjects on Entry into Cardiac Rehabilitation

Age (years) % Male Body mass index (kg/m2) Active smokers (%) Hypertensives (%) Diabetes (%) Married (%) Peak VO2 (mL/kg/min) Total cholesterol (mg/dL) HDL ⫺ cholesterol (mg/dL) LDL ⫺ cholesterol (mg/dL) Triglycerides (mg/dL) hs-CRP (mg/dL) Ejection fraction (%) Depression Anxiety Hostility Somatization Quality of life

801

Treatment (n ⫽ 91)

Control (n ⫽ 48)

P-value

62 ⫾ 12 70% 28.6 ⫾ 5.4 4% 34% 24% 78% 16.0 ⫾ 5.2 169 ⫾ 34 40 ⫾ 12 96 ⫾ 30 162 ⫾ 76 6.5 ⫾ 10.1 53 ⫾ 13 10.7 ⫾ 3.4 10.7 ⫾ 4.9 7.0 ⫾ 4.9 10.2 ⫾ 3.9 83.5 ⫾ 15.0

61 ⫾ 14 52% 29.4 ⫾ 5.9 6% 71% 26% 67% 13.5 ⫾ 3.9 176 ⫾ 39 45 ⫾ 16 96 ⫾ 29 189 ⫾ 153 10.3 ⫾ 19.8 54 ⫾ 14 11.3 ⫾ 3.6 11.6 ⫾ 4.5 5.8 ⫾ 5.4 11.0 ⫾ 4.1 77.9 ⫾ 17.1

NS .04 NS NS ⬍.0001 NS NS .004 NS .06 NS NS NS NS NS NS NS NS .06

All-cause mortality was assessed in control patients based on the presence or absence of depressive symptoms. In this cohort, depressed subjects had a nearly 3-fold higher mortality compared with nondepressed subjects (30% vs 11%, P ⫽ .003). All-cause mortality was further evaluated in the treatment cohort based on the presence or absence of depressive symptoms at completion of rehabilitation (Figure 2). In this cohort, depressed patients had more than a 4-fold higher mortality (22% vs 5%, P ⫽ .0004) than nondepressed patients. Time-dependent actuarial hazard for survival was assessed in depressed and nondepressed treatment patients (Figure 3). Depressed patients demonstrated an early and incremental increased mortality compared with nondepressed patients (P ⬍.001).

Table 4 Benefits of Cardiac Rehabilitations and Exercise Training in Depressed Coronary Patients (n ⫽ 91)

2

Body mass index (kg/m ) Peak VO2 (Ml/kg/min) Total cholesterol (mg/dL) HDL ⫺ cholesterol (mg/dL) LDL ⫺ cholesterol (mg/dL) Triglycerides (mg/dL) hs-CRP (mg/dL) Depression score Anxiety score Hostility score Somatization score Quality of life score

Before

After

P-value

28.6 ⫾ 5.4 16.0 ⫾ 5.2 169 ⫾ 34 40 ⫾ 12 96 ⫾ 30 162 ⫾ 76 6.4 ⫾ 10.1 10.7 ⫾ 3.4 10.7 ⫾ 4.9 7.0 ⫾ 4.9 10.2 ⫾ 3.9 83.4 ⫾ 15.0

28.3 ⫾ 5.3 17.9 ⫾ 5.8 169 ⫾ 33 43 ⫾ 11 101 ⫾ 41 147 ⫾ 95 4.8 ⫾ 6.8 3.9 ⫾ 4.2 4.5 ⫾ 4.3 3.4 ⫾ 4.2 6.4 ⫾ 4.1 107.4 ⫾ 17.7

NS ⬍.0001 NS .002 NS NS .04 ⬍.0001 ⬍.0001 ⬍.0001 ⬍.0001 ⬍.0001

802

The American Journal of Medicine, Vol 120, No 9, September 2007

Figure 1 Changes in prevalence of depression following cardiac rehabilitation and exercise training (n ⫽ 522).

By multivariate analysis, peak VO2, age, ejection fraction, and depressive symptoms remained independent predictors of death in the cardiac rehabilitation cohort (Table 5).

Impact of Cardiac Rehabilitation and Exercise Training on Depressive Symptoms and Subsequent Survival In order to evaluate the potential impact of rehabilitation on mortality in depressed subjects, we assessed mortality in treatment and control depressed patients identified at entry into cardiac rehabilitation (Figure 4). Depressed control patients had an almost 4-fold increase in mortality compared with depressed patients completing cardiac rehabilitation (30% vs 8%; P ⫽ .0005). Changes in depression status during the course of rehabilitation was obtained and evaluated as a function of subsequent mortality. Patients with depressive symptoms at

Figure 3 Actuarial cumulative hazard plot for survival time based on depression status upon completion of cardiac rehabilitation and exercise training (n ⫽ 522).

baseline who became nondepressed during the program appeared to have similar low mortality (6%), as did the 418 patients without depressive symptoms at either time (mortality 5%). In contrast, the 25 patients who had depressive symptoms at both baseline and postrehabilitation, as well as the 8 patients who did not meet criteria for depressive symptoms at baseline but did postrehabilitation, had high mortality (15% and 25%, respectively). Of the 25 patients, none of these patients developed interim cardiovascular events while enrolled in cardiac rehabilitation. In order to examine the singular impact of exercise training upon depressive symptoms, treatment patients were divided into 3 groups based on changes in peak VO2 during the course of cardiac rehabilitation (Figure 5). Those patients who did not improve functional status (exhibited as either no gain or a loss in peak VO2) were labeled “VO2 loss” (n ⫽ 102), those with a mild improvement (ⱕ10% increase in peak VO2) were labeled “mild VO2 gain” (n ⫽ 135), and those patients who increased peak VO2 by ⬎10% as “high VO2 gain” (n ⫽ 285). The prevalence of depressive symptoms was assessed in each of these groups at entry and completion of rehabilitation and subsequent mortality was obtained. There were no statistical differences in prevalence of depressive symptoms at baseline among the 3 groups. In the VO2 loss group, the prevalence of depressive symptoms remained unchanged (18% on entry to 14% on completion; P ⫽ NS) and the corresponding mortality

Table 5 Multivariate Analysis of Independent Predictors of Mortality in Patients who Completed Cardiac Rehabilitation (n ⫽ 522)

Figure 2 Mortality based on presence or absence of depressive symptoms upon completion of cardiac rehabilitation and exercise training (n ⫽ 522).

Variable

Chi-square

P-value

Peak VO2 Age Ejection fraction Depression

13.8 9.4 7.1 5.7

.0002 .002 .008 .017

Milani and Lavie

Exercise Reduces Depression-Related Mortality

Figure 4 Impact of cardiac rehabilitation and exercise training on mortality in 139 patients with baseline depression.

was relatively high at 15%. In patients with mild VO2 gain, the prevalence of depressive symptoms fell from 17% to 5% (P ⫽ .004) and the mortality was only 6%. Patients with high VO2 gain demonstrated similar improvements compared with those with only mild VO2 gain, with the prevalence of depressive symptoms decreasing from 18% to 5% (P ⬍.0001) accompanied by a mortality of only 4%.

DISCUSSION There are several important implications of this study. First, depressive symptoms are prevalent in the coronary heart disease population, particularly in patients following major coronary events. Second, the presence of depressive symptoms, regardless of age or sex, confers a marked increase in all-cause mortality risk over time. Third, cardiac rehabilitation is associated with a marked reduction in the prevalence of depressive symptoms corresponding to a significant improvement in survival. Finally, the full benefits of cardiac

803

rehabilitation on depressive symptoms and subsequent survival seem to be achieved by only a modest improvement in exercise capacity. Psychosocial factors, including depression, have been shown to be potent risk factors for the development of MI and death.9,10,27-29 The Interheart study evaluated cardiovascular risk factors in 29,972 people from 52 countries to investigate risk of first MI. Nine independent risk factors, including psychosocial factors, accounted for 90% of the population attributable risk in men and 94% in women. These psychosocial factors, which included depression, increased the odds of first MI by nearly 3-fold and accounted for 33% of the population attributable risk for the development of MI, a magnitude similar to standard risk factors such as smoking, diabetes and hypertension.30 Within this psychosocial index, depression alone increased the odds of MI by nearly 2-fold.31 Following a major cardiac event, the risk of subsequent MI and death related to depression appears much higher.9,32-34 Moreover, several studies have documented the poor recovery and durability of untreated depression following MI.4,13,35 As a result, there is strong interest in determining whether prognosis can be improved in cardiac patients with depression. The ENRICHD (Enhancing Recovery in Coronary Heart Disease Patients) trial randomized depressed post-MI patients to cognitive behavior therapy and after an average follow-up of 29 months found improvement in depression and social isolation without any change in survival.36 However, the ENRICHD investigators noted the association between self-reported exercise following MI and reduced depression and improved survival.37 Carney et al have shown that patients who remained depressed despite treatment had worse outcomes than those patients who were either not depressed at baseline or who become less depressed following treatment.38 Data from a recent randomized trial demonstrated that exercise training can decrease depressive symptoms as effectively as antidepressant med-

Figure 5 Prevalence of depression and subsequent mortality based on changes in peak oxygen consumption (VO2) during cardiac rehabilitation and exercise training. *P ⬍.001 compared with VO2 loss.

804

The American Journal of Medicine, Vol 120, No 9, September 2007

ications.20,21 To our knowledge, our investigation is the first to demonstrate a survival benefit resulting from reducing depressive symptoms and improving fitness in cardiac patients. We and others have demonstrated the marked benefits of cardiac rehabilitation and exercise training on conventional as well as psychological factors,15,16,39-41 with data from meta-analyses and other trials demonstrating improvements in overall mortality.42,43 Whether or not this survival benefit comes from improvement in exercise capacity or some other mechanism has been uncertain. Previously, Vanhees et al demonstrated that peak VO2 evaluated after physical training and its change in response to physical training are independent predictors of cardiovascular mortality, such that every 1% increase in peak VO2 results in a 2% decrease in cardiovascular death.44 Our data support the importance of cardiac rehabilitation in improving exercise capacity but also in reducing symptoms of depression and its accompanying high mortality. The impact of cardiac rehabilitation upon depressive symptoms is likely multi-factorial. Cardiac rehabilitation is centered around progressive exercise training, which has been shown to exert a salutary effect on certain emotions.24,45-48 In a review of the literature, Byrne and Byrne reported a favorable trend between exercise status and mood states, particularly depression.49 The Alameda County Study used questionnaires to measure depressive symptoms and physical activity in 8023 adults over 3 time periods, beginning in 1965. Individuals with low physical activity were at significantly greater risk for depression at a 9-year follow-up evaluation. Additionally, those individuals who had been inactive in 1965 but increased their activity levels by 1974 were at no greater risk for depression in 1983 than those who had been highly active all along, suggesting “the high risk of depression associated with inactivity is modified if the activity level is changed.”50 Recent data also emphasize the benefits of exercise on cognitive function and brain plasticity.51 In the current investigation, we attempted to isolate the exercise component of cardiac rehabilitation by grouping subjects based on changes in peak VO2 over the course of the program. Patients with no improvement in exercise capacity attained no significant reduction in depressive symptoms, whereas those who achieved a modest (1%10% increase in peak VO2) or more robust (⬎10% increase in peak VO2) enhancement in exercise capacity demonstrated more than a 70% reduction in the prevalence of depressive symptoms. Moreover, survival mirrored reduction in depressive symptoms, with the best outcomes occurring in groups who improved peak VO2. This suggests not only a favorable relationship between exercise training and depressive symptoms, but also that only a modest enhancement in overall fitness is required to attain the benefits of this intervention on both this behavior and subsequent survival. Another possible mechanism of reducing depressive symptoms in cardiac rehabilitation is the education of the

patient and spouse. The program includes educational sessions on coronary heart disease and risk factor modification, with frequent opportunities for addressing patient/spouse concerns, thus empowering the patient to modify their own recovery. This results in a mechanism by which patients become involved in their own health care, coined “information involvement,” which has been shown to influence coping and social-emotional recovery following major cardiac events.52,53 Socialization and bonding with other patients, who are at various stages of recovery and rehabilitation, is an important component of the intervention program and likely contributes to the favorable effects seen upon depression.15,54,55 Finally, dietary changes, including increased consumption of foods rich in n-3 fatty acids as part of the Mediterranean diet, are part of our recommendations, and deficiencies in this important nutrient have been correlated to depression, although our study did not allow for analysis of this factor.56,57 Although a limitation to this study was the lack of a randomized nonrehabilitation control group where depressive symptoms were assessed at 2 points in time (which makes establishing cause and effect more difficult), we have previously demonstrated lack of improvement in most coronary risk factors, including behavioral factors, in control subjects not entering cardiac rehabilitation following major coronary heart disease events.26 Moreover, data specifically examining depression in coronary patients over time suggest that the disorder is fairly resistant and likely to remain if left untreated.13,35 Finally, cardiac rehabilitation is a clinically proven modality for reducing morbidity and mortality in coronary patients, and a randomized control group not receiving this therapy would be medically unethical.39,40,42,43 Another study limitation may be selection bias, because the study group did not represent all patients who have coronary events but rather those selected for referral and who attended and completed the formal cardiac rehabilitation program. In addition, the instrument used to assess depressive symptoms is not as well established as some other instruments and has not been demonstrated previously to predict adverse events, nor has it been used in clinical trials of depression. Nevertheless, our incidence rate of depressive symptoms upon entry into cardiac rehabilitation was very similar to that of other investigators, and suggests that our sample was representative of patients following acute cardiac events. More importantly, depressive symptoms demonstrated in this study were associated with markedly increased mortality in both patients not completing rehabilitation and those patients with depressive symptoms following cardiac rehabilitation, thus validating the prognostic impact of this specific assessment. Finally, our study assessed all-cause mortality, a very powerful endpoint. Although our depressed coronary patients did not have interim events during cardiac rehabilitation, our data do not allow us to determine the impact of cardiac rehabilitation and depression on other potentially important endpoints.

Milani and Lavie

Exercise Reduces Depression-Related Mortality

In conclusion, depressed patients following major coronary events have a significantly higher mortality than nondepressed patients. Depressive symptoms and the excess mortality associated with them can be substantially reduced utilizing cardiac rehabilitation and exercise training. Exercise therapy and its associated improvement in fitness appears to be the major mechanism affecting depressive symptoms and subsequent mortality.

References 1. Carney RM, Rich MW, Tevelde A, et al. Major depressive disorder in coronary artery disease. Am J Cardiol. 1987;60(16):1273-1275. 2. Schleifer SJ, Macari-Hinson MM, Coyle DA, et al. The nature and course of depression following myocardial infarction. Arch Intern Med. 1989;149(8):1785-1789. 3. Forrester AW, Lipsey JR, Teitelbaum ML, et al. Depression following myocardial infarction. Int J Psychiatry Med. 1992;22(1):33-46. 4. Bush DE, Ziegelstein RC, Patel UV, et al. Post-myocardial infarction depression. Evid Rep Technol Assess (Summ). 2005(123):1-8. 5. Rumsfeld JS, Ho PM. Depression and cardiovascular disease: a call for recognition. Circulation. 2005;111(3):250-253. 6. Frasure-Smith N, Lesperance F. Reflections on depression as a cardiac risk factor. Psychosom Med. 2005;67(Suppl 1):S19-S25. 7. Rozanski A, Blumenthal JA, Davidson KW, et al. The epidemiology, pathophysiology, and management of psychosocial risk factors in cardiac practice: the emerging field of behavioral cardiology. J Am Coll Cardiol. 2005;45(5):637-651. 8. Carney RM, Rich MW, Freedland KE, et al. Major depressive disorder predicts cardiac events in patients with coronary artery disease. Psychosom Med. 1988;50(6):627-633. 9. Frasure-Smith N, Lesperance F, Talajic M. Depression following myocardial infarction. Impact on 6-month survival. JAMA. 1993;270(15): 1819-1825. 10. Frasure-Smith N, Lesperance F, Talajic M. Depression and 18-month prognosis after myocardial infarction. Circulation. 1995;91(4):9991005. 11. van Melle JP, de Jonge P, Spijkerman TA, et al. Prognostic association of depression following myocardial infarction with mortality and cardiovascular events: a meta-analysis. Psychosom Med. 2004;66(6):814822. 12. Ho PM, Masoudi FA, Spertus JA, et al. Depression predicts mortality following cardiac valve surgery. Ann Thorac Surg. 2005;79(4):12551259. 13. Ladwig KH, Roll G, Breithardt G, et al. Post-infarction depression and incomplete recovery 6 months after acute myocardial infarction. Lancet. 1994;343(8888):20-23. 14. Milani RV, Lavie CJ, Cassidy MM. Effects of cardiac rehabilitation and exercise training programs on depression in patients after major coronary events. Am Heart J. 1996;132(4):726-732. 15. Milani RV, Lavie CJ. Prevalence and effects of cardiac rehabilitation on depression in the elderly with coronary heart disease. Am J Cardiol. 1998;81(10):1233-1236. 16. Lavie CJ, Milani RV, Cassidy MM, Gilliland YE. Effects of cardiac rehabilitation and exercise training programs in women with depression. Am J Cardiol. 1999;83(10):1480-1483, A1487. 17. Lavie CJ, Milani RV. Cardiac rehabilitation and depression. Am J Cardiol. 2004;93(8):1080. 18. Blumenthal JA, Sherwood A, Babyak MA, et al. Effects of exercise and stress management training on markers of cardiovascular risk in patients with ischemic heart disease: a randomized controlled trial. JAMA. 2005;293(13):1626-1634. 19. Lavie CJ, Milani RV. Adverse psychological and coronary risk profiles in young patients with coronary artery disease and benefits of formal cardiac rehabilitation. Arch Intern Med. 2006;166:1878-1883.

805

20. Blumenthal JA, Babyak MA, Moore KA, et al. Effects of exercise training on older patients with major depression. Arch Intern Med. 1999;159(19):2349-2356. 21. Babyak M, Blumenthal JA, Herman S, et al. Exercise treatment for major depression: maintenance of therapeutic benefit at 10 months. Psychosom Med. 2000;62(5):633-638. 22. Kellner R. A symptom questionnaire. J Clin Psychiatry. 1987;48(7): 268-274. 23. Stewart AL, Greenfield S, Hays RD, et al. Functional status and well-being of patients with chronic conditions. Results from the Medical Outcomes Study. JAMA. 1989;262(7):907-913. 24. Lavie CJ, Milani RV. Benefits of cardiac rehabilitation and exercise training. Chest. 2000;117(1):5-7. 25. Milani RV, Lavie CJ, Mehra MR, Ventura HO. Understanding the basics of cardiopulmonary exercise testing. Mayo Clin Proc. 2006; 81(12):1603-1611. 26. Milani RV, Lavie CJ, Mehra MR. Reduction in C-reactive protein through cardiac rehabilitation and exercise training. J Am Coll Cardiol. 2004;43(6):1056-1061. 27. Barefoot JC, Helms MJ, Mark DB, et al. Depression and long-term mortality risk in patients with coronary artery disease. Am J Cardiol. 1996;78(6):613-617. 28. Lesperance F, Frasure-Smith N, Juneau M, Theroux P. Depression and 1-year prognosis in unstable angina. Arch Intern Med. 2000;160(9): 1354-1360. 29. Lesperance F, Frasure-Smith N, Talajic M, Bourassa MG. Five-year risk of cardiac mortality in relation to initial severity and one-year changes in depression symptoms after myocardial infarction. Circulation. 2002;105(9):1049-1053. 30. Yusuf S, Hawken S, Ounpuu S, et al. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control study. Lancet. 2004;364(9438): 937-952. 31. Rosengren A, Hawken S, Ounpuu S, et al. Association of psychosocial risk factors with risk of acute myocardial infarction in 11119 cases and 13648 controls from 52 countries (the INTERHEART study): casecontrol study. Lancet. 2004;364(9438):953-962. 32. Blumenthal JA, Lett HS, Babyak MA, et al. Depression as a risk factor for mortality after coronary artery bypass surgery. Lancet. 2003; 362(9384):604-609. 33. Bush DE, Ziegelstein RC, Tayback M, et al. Even minimal symptoms of depression increase mortality risk after acute myocardial infarction. Am J Cardiol. 2001;88(4):337-341. 34. Carney RM, Blumenthal JA, Catellier D, et al. Depression as a risk factor for mortality after acute myocardial infarction. Am J Cardiol. 2003;92(11):1277-1281. 35. Stern MJ, Pascale L, Ackerman A. Life adjustment postmyocardial infarction: determining predictive variables. Arch Intern Med. 1977; 137(12):1680-1685. 36. Berkman LF, Blumenthal J, Burg M, et al. Effects of treating depression and low perceived social support on clinical events after myocardial infarction: the Enhancing Recovery in Coronary Heart Disease Patients (ENRICHD) Randomized Trial. JAMA. 2003;289(23):31063116. 37. Blumenthal JA, Babyak MA, Carney RM, et al. Exercise, depression, and mortality after myocardial infarction in the ENRICHD trial. Med Sci Sports Exerc. 2004;36(5):746-755. 38. Carney RM, Blumenthal JA, Freedland KE, et al. Depression and late mortality after myocardial infarction in the Enhancing Recovery in Coronary Heart Disease (ENRICHD) study. Psychosom Med. 2004; 66(4):466-474. 39. Wenger NK, Froelicher ES, Smith LK, et al. Cardiac rehabilitation as secondary prevention. Agency for Health Care Policy and Research and National Heart, Lung, and Blood Institute. Clin Pract Guidel Quick Ref Guide Clin. 1995(17):1-23. 40. Ades PA. Cardiac rehabilitation and secondary prevention of coronary heart disease. N Engl J Med. 2001;345(12):892-902.

806

The American Journal of Medicine, Vol 120, No 9, September 2007

41. Lavie CJ, Milani RV, Littman AB. Benefits of cardiac rehabilitation and exercise training in secondary coronary prevention in the elderly. J Am Coll Cardiol. 1993;22(3):678-683. 42. O’Connor GT, Buring JE, Yusuf S, et al. An overview of randomized trials of rehabilitation with exercise after myocardial infarction. Circulation. 1989;80(2):234-244. 43. Witt BJ, Jacobsen SJ, Weston SA, et al. Cardiac rehabilitation after myocardial infarction in the community. J Am Coll Cardiol. 2004; 44(5):988-996. 44. Vanhees L, Fagard R, Thijs L, Amery A. Prognostic value of traininginduced change in peak exercise capacity in patients with myocardial infarcts and patients with coronary bypass surgery. Am J Cardiol. 1995;76(14):1014-1019. 45. Blumenthal JA, Fredrikson M, Kuhn CM, et al. Aerobic exercise reduces levels of cardiovascular and sympathoadrenal responses to mental stress in subjects without prior evidence of myocardial ischemia. Am J Cardiol. 1990;65(1):93-98. 46. Blumenthal JA, Williams RS, Needels TL, Wallace AG. Psychological changes accompany aerobic exercise in healthy middle-aged adults. Psychosom Med. 1982;44(6):529-536. 47. Stern MJ, Cleary P. National Exercise and Heart Disease Project. Psychosocial changes observed during a low-level exercise program. Arch Intern Med. 1981;141(11):1463-1467. 48. Milani RV, Lavie CJ. Behavioral differences and effects of cardiac rehabilitation in diabetic patients following cardiac events. Am J Med. 1996;100(5):517-523.

49. Byrne A, Byrne DG. The effect of exercise on depression, anxiety and other mood states: a review. J Psychosom Res. 1993;37(6):565-574. 50. Camacho TC, Roberts RE, Lazarus NB, et al. Physical activity and depression: evidence from the Alameda County Study. Am J Epidemiol. 1991;134(2):220-231. 51. McAuley E, Kramer AF, Colcombe SJ. Cardiovascular fitness and neurocognitive function in older adults: a brief review. Brain Behav Immun. 2004;18(3):214-220. 52. Krantz DS, Baum A, Wideman M. Assessment of Preferences for self-treatment and information in health care. J Pers Soc Psychol. 1980;39(5):977-990. 53. Mahler HI, Kulik JA. Health care involvement preferences and socialemotional recovery of male coronary-artery-bypass patients. Health Psychol. 1991;10(6):399-408. 54. Kulik JA, Mahler HI. Social support and recovery from surgery. Health Psychol. 1989;8(2):221-238. 55. Cohen F, Lazarus RS. Active coping processes, coping dispositions, and recovery from surgery. Psychosom Med. 1973;35(5):375-389. 56. de Lorgeril M, Salen P, Martin JL, et al. Mediterranean diet, traditional risk factors, and the rate of cardiovascular complications after myocardial infarction: final report of the Lyon Diet Heart Study. Circulation. 1999;99(6):779-785. 57. Severus WE, Littman AB, Stoll AL. Omega-3 fatty acids, homocysteine, and the increased risk of cardiovascular mortality in major depressive disorder. Harv Rev Psychiatry. 2001;9(6):280-293.

The American Journal of Medicine (2007) 120, 807-813

CLINICAL RESEARCH STUDY

Independent Vascular and Cognitive Risk Factors for Postoperative Delirium James L. Rudolph, MD, SM,a,h,i Richard N. Jones, ScD,b,e,i Lars S. Rasmussen, MD,c Jeffrey H. Silverstein, MD,d Sharon K. Inouye, MD, MPH, b,e,i Edward R. Marcantonio, MD, SMb,e,f,i a

Geriatric Research, Education, and Clinical Center, VA Boston Healthcare System, Boston, Mass; bInstitute for Aging Research and Aging Brain Center, Hebrew SeniorLife, Boston, Mass; cCentre of Head and Orthopaedics, Department of Anaesthesia, Copenhagen University Hospital, Copenhagen, Denmark; dDepartments of Anesthesiology, Surgery, and Geriatrics and Adult Development, Mount Sinai School of Medicine, New York, NY; eDivisions of Gerontology and fGeneral Medicine and Primary Care, Beth Israel Deaconess Medical Center, Boston, Mass; hDivision of Aging, Brigham and Women’s Hospital, Boston, Mass; iDepartment of Medicine, Harvard Medical School, Boston Mass. ABSTRACT BACKGROUND: Delirium is a common, morbid, and costly syndrome that occurs frequently after surgery for atherosclerosis. We hypothesized that vascular risk factors and mildly impaired cognitive performance would independently predispose nondemented patients to develop delirium after noncardiac surgery. METHODS: The International Study of Postoperative Cognitive Dysfunction recruited patients undergoing noncardiac surgery from 8 countries. Subjects provided detailed medical history and underwent preoperative testing of multiple cognitive domains with a neuropsychologic battery. Postoperatively, subjects (n ⫽ 1161) were assessed daily for delirium. RESULTS: Ninety-nine subjects (8%) developed delirium. In bivariable analysis, several vascular risk factors were significantly associated with the likelihood of delirium, including male sex, exposure to tobacco, previous myocardial infarction, and vascular surgery. After adjustment for age, tobacco exposure and vascular surgery were independent vascular risk factors for delirium (adjusted relative risk [RR] 3.2, 95% confidence interval [CI], 2.1-4.9). In addition, mildly impaired cognitive performance, defined as performance 1.5 standard deviation below the mean on either of 2 neuropsychologic tests, was independently associated with delirium (adjusted RR 2.2, 95% CI, 1.4-3.6). Subjects with both vascular risk factors and mildly impaired cognitive performance were at double the risk of delirium (RR 2.2, 95% CI, 1.2-4.2) compared with those with either of these risk factors alone. CONCLUSIONS: Vascular risk and mildly impaired cognitive performance independently predispose patients to delirium after noncardiac surgery. These factors will help to identify high-risk patients for delirium and to design and target future intervention strategies. © 2007 Elsevier Inc. All rights reserved. KEYWORDS: Atherosclerosis; Cognition; Delirium; Risk factors; Surgery; Tobacco; Vascular surgery

Delirium is defined as an acute disorder of attention and cognition with high rates of adverse outcomes. For instance, the in-hospital mortality of postoperative delirium is 4% to 13%,1,2 similar to that of acute myocardial infarction or These data were collected as a result of funding from multiple sources.16 The authors were supported in part by National Institute of Health grants: 5 K12 AG00294-18 (Dr Rudolph), 5 P60 AG08812-14 (Drs Marcantonio and Jones), RO1AG018772-04 (Dr Silverstein), and R21AG025193 and K24AG000949 (Dr Inouye). Dr Rasmussen is

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2007.02.026

sepsis.3 In addition, delirium in the surgical population is associated with increased costs,4 prolonged length of stay, increased rate of discharge to a nursing home, increased supported in part by ongoing research studies on postoperative complications. Dr Marcantonio is a Paul Beeson Physician Faculty Scholar in Aging Research. Requests for reprints should be addressed to James L. Rudolph, MD, VABHS (JP-182), 150 South Huntington Ave, Boston, MA 02130. E-mail address:[email protected].

808

The American Journal of Medicine, Vol 120, No 9, September 2007

functional impairment 6 months after surgery,5 and inlanguage of the testing, emergency surgery, outpatient adcreased 1-year mortality.6 Despite its clinical importance, ministration of tranquilizers or antidepressants, neurosurdelirium is often unrecognized by physicians.3,7 gery, severe hearing or vision disorders, life expectancy less Atherosclerosis may predispose patients to delirium. Within than 3 months, or refusal. The protocol was approved by the the surgical population, the incidence of delirium after surgery research ethics committees at all centers, and all subjects for atherosclerosis, such as aortic provided written informed conaneurysm repair (41%-54%), corosent. Of the 1218 subjects ennary artery bypass grafting (32%rolled, 57 (5%) did not have postCLINICAL SIGNIFICANCE 50%), and peripheral vascular byoperative delirium assessments. pass (10%-48%), is approximately Subjects with and without delir● Delirium after noncardiac surgery is double that of elective orthopedic ium assessments had similar basecommon and associated with increased (9%-15%), abdominal (5%-26%), line characteristics and neuropsymortality. and head and neck procedures chologic performance (P ⬎ .05). ● Vascular risk factors such as age, to(17%).8 In addition, previous studThe exception was the MMSE; ies have identified atherosclerosis those missing delirium assessbacco use, and the need for vascular burden9 and vascular risk factors ments scored better (28.2 vs 27.8; surgery were independently associated such as advanced age,1,3 male genP ⫽ .05). with postoperative delirium. der,3 tobacco use,10,11 and hyperten● In patients without dementia and Minision10 as important risk factors for Vascular Risk Factors Mental State Examination scores of 24 or delirium. Atherosclerosis is an imBefore surgery, all subjects undermore, mildly impaired cognitive perforportant contributor to a spectrum of went extensive interviews and cognitive impairments, including mance was independently associated medical record reviews. Demodementia. Although dementia is with delirium. graphic data known to predispose clearly a risk factor for delirium,12 to atherosclerosis, such as age and ● Patients with vascular or cognitive risk less is known about milder forms of gender, were collected from subfactors for delirium may benefit from cognitive impairment and the risk of jects. Subjects provided informapreoperative screening and intervendelirium. Poor cognitive perfortion on alcohol consumption tions to prevent delirium. mance has been identified as a risk (drinks per week) and tobacco use factor for both dementia13 and (pack-years). From the medical delirium.14,15 record, we collected a history of The aims of this study were to conditions known to predispose a examine vascular risk factors and determine their indepenperson to vascular risk, including hypertension, diabetes, dent contributions to delirium; examine mildly impaired previous myocardial infarction, congestive heart failure, ancognitive performance on neuropsychologic tests and detergina, the need for vascular surgery, and so forth. mine its independent contribution to delirium; and examine the interrelationship between vascular risk factors and Neuropsychologic Assessment mildly impaired cognitive performance in their contribuA battery of neuropsychologic tests were completed on tions to delirium risk. We hypothesized that both vascular entry into the study, generally the day before surgery, to risk factors and mildly impaired cognitive performance assess multiple cognitive domains. The battery was selected would independently contribute to delirium. because it was sensitive, not influenced by culture, and completed in 45 minutes or less.16 Trained interviewers, METHODS blinded to study hypotheses, administered the tests. Regular site visits were made to ensure uniform test administration Recruitment and data collection.16 The International Study of Postoperative Cognitive DysThe battery was derived from standard neuropsychologic function (ISPOCD) recruited 1218 patients aged 60 years or tests, translated into the participating country’s native lanmore at 13 hospitals in 8 countries (Denmark, France, Gerguage, and validated in a separate population.16 The battery many, Great Britain, Greece, The Netherlands, Spain, and has been described in detail18 and consists of the following United States) undergoing noncardiac surgery between Nomeasures. The Visual Verbal Learning Test (VVLT), based vember 1, 1994, and May 31, 1996. This study has been on Rey’s auditory recall of words,19 consists of 15 words described in detail.16 Eligible surgical procedures included presented at a fixed rate. At the end of the 3 learning trials orthopedic, abdominal, thoracic, vascular, or other noncarand a 20-minute delay, subjects were asked to recall as diac procedures with anticipated general anesthesia and an many words as possible. We examined the number of words inpatient stay. Patients were excluded if they had a score of recalled over the 3 learning trials (VVLT Learning Score) 23 or less on the Mini-Mental State Examination and the percentage of words recalled after a delay (VVLT (MMSE),17 dementia, Parkinson disease, previous neuroMemory Percent). The Concept Shifting Test and the Stroop psychologic testing, illiteracy, inability to understand the

Rudolph et al

Vascular Risk Factors for Delirium

Color-Word interference test are timed measures that consist of 2 priming trials followed by an interference trial.16,18 The Memory Scanning Test20 is a timed test that presented 1-, 2-, and 3-letter combinations and, from memory, asked subjects to identify the letter(s) among a field of distracting information. The time to complete and errors for all timed tests were selected for analysis. The number of correct answers was recorded from the Letter-Digit Substitution, derived from the Symbol-Digit Substitution Test.21 The Four Boxes Test22 measured reaction time and errors over 52 responses. The number of correct items from the Cattell Culture Fair Intelligence Test was recorded.23 Neuropsychologic scores were standardized to the overall mean and standard deviation (SD) (z scores) and reversed as appropriate for timed and error assessments to provide consistent analysis (lower score indicating worse performance). For each neuropsychologic measure, subjects having mildly impaired cognitive performance were identified as those performing 1.5 SD or less below the mean. This threshold is consistent with clinically meaningful risk categories.13 Because all patients with overt dementia or MMSE scores less than 24 were excluded, there was no lower limit to the mildly impaired cognitive performance definition.

Delirium Beginning on the operative day, a trained interviewer administered the orientation questions of the MMSE daily until the earlier of either hospital discharge or postoperative day (POD) 7. The presence or absence of delirium according to the criteria of the Diagnostic and Statistical Manual of Mental Disorders, 3rd Edition (DSM-III) was recorded.24 For the time-to-event analysis, we identified the POD that the subject first met criteria for delirium. The outcome of this study was the presence of delirium on any day after surgery.

Background Variables Age has been demonstrated to be a risk factor for cognitive impairment,25 atherosclerosis,26 and delirium,1,14 and was included as an adjustment variable in multivariable analyses. Subjects with any tobacco use were compared with those with no exposure.27 Subjects who consumed 1 to 14 alcoholic drinks per week were compared with nondrinkers28 and heavy drinkers (⬎14 drinks per week). There was no uniform study protocol to identify and treat alcohol or tobacco withdrawal symptoms. Medical history related to vascular risk and/or disease was recorded as dichotomous.

Data Analysis All data were entered into a database at each site and sent to the data coordinating site electronically. The coordinating site (Copenhagen) and the data management site (Eidenhoven) performed error and range checking of the data. The system was checked for reliability during a pilot study in all centers.16 We performed our statistical analyses with the

809 Statistical Package for the Social Sciences version 11.5.0 (SPSS, Inc, Chicago, Ill). Our bivariable analysis compared the risk of delirium when the vascular risk factor was present with the rate when the risk factor was absent. Vascular risk factors with a relative risk (RR) of 1.5 or more in the bivariable analysis were entered into stepwise backward and forward proportional hazard regression models requiring a P value less than .10 to enter the model and a P value less than .05 to remain in the model. Age was included as a covariate in the stepwise regression models. Proportional hazards models were used because the models compare counts (rates) among those with and without delirium, make no assumptions about the underlying data, and directly generate a RR. Fulfillment of the proportionality assumption was verified for all analyses by testing the interaction between independent variables and the time-to-event outcome. Vascular risk factors that remained in the final model were considered independent predictors of delirium. To determine the bivariable risk of delirium, we compared the incidence of postoperative delirium in subjects with and without mildly impaired cognitive performance on each neuropsychologic measure. Neuropsychologic measures with a bivariable RR of 1.5 or more were entered into backward and forward proportional hazard models that included age. The neuropsychologic measures that remained in the final model were considered independent risk factors for delirium. As a final step, to determine whether vascular risk factors were independent of mildly impaired cognitive performance, we entered the independent vascular risk factor count and mildly impaired cognitive performance into a combined proportional hazards model that included age adjustment.

RESULTS The characteristics of the sample are summarized in Table 1. The resultant sample had a mean age of 68.9 (SD ⫾ 5.9) years and a slight preponderance of males (53%). The mean preoperative MMSE score was 27.8 (⫾1.6). The cohort included subjects with vascular risk factors, such as a history of tobacco use in 685 (59%), diabetes in 86 (7%), hypertension in 398 (34%), previous myocardial infarction in 102 (9%), congestive heart failure in 132 (11%), and angina pectoris in 100 (9%). The sample underwent a spectrum of surgical procedures, including abdominal (35%), genitourinary (21%), vascular (15%), orthopedic (15%), thoracic (3%), and other procedures (10%). Completion of baseline cognitive instruments ranged from 1081 (93%) for the Four Boxes Test to 1155 (99%) for the VVLT Learning Score. The cumulative incidence of delirium in the 7-day postoperative period was 99 of 1161 (9%), with a peak incidence on POD 2 (5%). In those who developed delirium on POD 1, 74% had symptoms of delirium on subsequent days. Subjects with delirium had a higher inpatient mortality

810

The American Journal of Medicine, Vol 120, No 9, September 2007

Table 1

Baseline Characteristics Mean ⫾ SD n (%)

Characteristic

N

Age (y) Male n (%) Tobacco (pack-y) Alcohol (drinks/wk) Functional impairment n (%) Mini-Mental State Examination Hypertension n (%) Diabetes n (%) Congestive heart failure n (%) Previous myocardial infarction n (%) Angina n (%) Stroke n (%) Peripheral vascular disease n (%) Surgery type Abdominal n (%) Genitourinary n (%) Vascular n (%) Orthopedic n (%) Other (eg, skin, breast) n (%) Thoracic n (%)

1161 1161 1160 1160 1159 1154 1159 1159 1161 1159 1159 1159 1159 1159

68.9 618 17.1 5.4 205 27.8 398 86 132 102 100 32 189

⫾ 5.9 (53.2) ⫾ 21.3 ⫾ 10.2 (17.7) ⫾ 1.6 (34.3) (7.4) (11.4) (8.8) (8.6) (2.8) (16.3)

411 245 175 171 117 40

(35.5) (21.1) (15.1) (14.8) (10.1) (3.5)

SD ⫽ standard deviation.

compared with those who did not develop delirium (8% vs 2%, P ⬍ .001). In addition, subjects with delirium were more likely to have an increased hospital length-of-stay compared with those without delirium (18 ⫾ 13 days vs 12 ⫾ 16 days, P ⬍ .001). Table 2 describes the incidence of delirium and RR associated with the individual vascular risk factors. In

Table 2

bivariable analysis, the risk of delirium was significantly increased in males, subjects with a history of tobacco use or myocardial infarction, and those who underwent vascular surgery. In the age-adjusted stepwise model, tobacco exposure (adjusted RR 1.6, 95% confidence interval [CI], 1.0-2.6) and vascular surgery (adjusted RR 2.7, 95% CI, 1.7-4.2) were the independent risk factors for delirium. Subjects with neither tobacco exposure nor vascular surgery developed delirium at a similar rate to those who had 1 vascular risk factor (5.4% vs 7.6%, P ⫽ .18). These groups were combined and compared with subjects with both independent vascular risk factors. Table 3 presents the incidence and RR of delirium among subjects with mildly impaired cognitive performance compared with those with better performance. The crude analysis demonstrates several neuropsychologic measures for which mildly impaired cognitive performance was associated with delirium. In the age-adjusted stepwise regression analyses, mildly impaired cognitive performance on the VVLT Learning Score (adjusted RR 2.0, 95% CI, 1.13.9) and errors on the Stroop Color-Word test (adjusted RR 2.3, 95% CI, 1.3-4.1) were independently associated with the risk of delirium. The rate of delirium was similar with mildly impaired cognitive performance on 1 independent test compared with 2 (21% vs 33%, P ⫽ .31). Thus, we designated mildly impaired cognitive performance as performance of 1.5 SD or more below the mean on either of these 2 tests. Table 4 describes the risk for delirium in subjects stratified for both vascular risk factors and mildly impaired cognitive performance. In subjects without mildly impaired

Vascular Risk Factors for Delirium Delirium When Risk Factor:

Vascular Risk Factor

Present n/N (%)

Absent n/N (%)

Crude RR (95% CI)

Adjusted RR† (95% CI)

Age (per y) Male Any tobacco exposure No alcohol use‡ Functional impairment Hypertension Diabetes Congestive heart failure Previous myocardial infarct Angina Stroke Peripheral vascular disease Vascular surgery§

32/543 27/476 52/618 81/954 68/761 87/1073 84/1029 83/1057 88/1059 95/1127 79/970 64/984

1.1 1.8 1.8 1.0 1.0 0.9 1.7 1.4 2.0 1.3 1.5 1.3 3.1

1.1 (1.0-1.1)

67/618 72/685 35/413 18/205 31/398 12/86 15/132 16/102 11/100 4/32 20/189 35/175

(10.8) (10.5) (8.5) (8.8) (7.8) (14.0) (11.4) (15.7) (11.0) (12.5) (10.6) (20.0)

(5.9) (5.7) (8.4) (8.5) (8.9) (8.1) (8.2) (7.9) (8.3) (8.4) (8.1) (6.5)

(1.0-1.1) (1.2-2.8)* (1.2-2.8)* (0.7-1.5) (0.6-1.7) (0.6-1.3) (0.9-3.0)* (0.8-2.3) (1.2-3.3)* (0.7-2.4) (0.6-3.8) (0.8-2.1) (2.1-4.5)*

1.6 (1.0-2.6)

2.7 (1.7-4.2)

RR ⫽ relative risk; CI ⫽ confidence interval; n ⫽ number of subjects with delirium; N ⫽ number of subjects with risk factor. *Represents variables with unadjusted RR more than 1.5 that were selected for entry into the forward and backward stepwise proportional hazards models. †Results of the stepwise proportional hazards model with adjustment for age. ‡Subjects who consumed alcohol in excess of 14 drinks per week developed delirium at a similar rate (12/129 [9.3%]) to nonconsumers and modest consumers. §Vascular surgery was compared with all other surgical procedures. Other surgical procedures (incidence of delirium) included abdominal (9.7%), orthopedic (5.3%), genitourinary (2.9%), thoracic (7.5%), and other (3.4%).

Rudolph et al Table 3

Vascular Risk Factors for Delirium

811

Mildly Impaired Cognitive Performance and Risk for Delirium Delirium When Performance ⱕ 1.5 SD Below Mean:

Neuropsychologic Measure

Present n/N (%)

Absent n/N (%)

Crude RR (95% CI)

Adjusted RR† (95% CI)

Age (per y) VVLT Learning Score VVLT Memory Percent 1-letter Memory Scanning Test 1-letter Memory Scanning Test 2-letter Memory Scanning Test 2-letter Memory Scanning Test 3-letter Memory Scanning Test 3-letter Memory Scanning Test Stroop Color-Word time Stroop Color-Word errors Concept Shifting Test time Concept Shifting Test errors Letter Digit Substitution Culture Fair Intelligence Test Four Boxes Test mean time Four Box Tests errors

14/59 13/72 15/86 4/62 17/72 6/75 11/74 10/73 13/75 16/72 11/80 16/118 10/90 5/69 18/102 4/35

85/1096 (7.8) 86/1083 (7.9) 84/1062 (7.9) 95/1088 (8.7) 84/978 (7.9) 93/976 (8.7) 88/1062 (8.3) 89/1065 (8.4) 83/1069 (7.8) 80/1072 (7.5) 87/1051 (8.3) 83/1016 (8.2) 89/1065 (8.4) 91/1061 (8.6) 75/979 (7.7) 89/1061(8.4)

1.1 3.1 2.3 2.2 0.7 2.2 0.8 1.6 1.4 2.2 3.0 1.7 1.7 1.3 0.8 2.3 1.4

1.1 (1.0-1.1) 2.0 (1.1-3.9)

time errors time errors time errors

(23.7) (18.1) (17.4) (6.5) (17.2) (7.4) (12.9) (12.0) (17.3) (22.2) (13.8) (13.6) (11.1) (7.2) (17.6) (11.4)

(1.0-1.1) (1.8-5.0)* (1.3-3.9)* (1.3-3.6)* (0.3-1.9) (1.3-3.6)* (0.4-1.8)* (0.9-2.8)* (0.8-2.7) (1.3-3.8)* (1.8-4.8)* (0.9-3.0)* (1.0-2.7)* (0.7-2.5) (0.4-2.0) (1.4-3.7)* (0.5-3.5)

2.3 (1.2-4.1)

SD ⫽ standard deviation; RR ⫽ relative risk; CI ⫽ confidence interval; VVLT ⫽ Visual Verbal Learning Test; n ⫽ number of subjects with delirium; N ⫽ number of subjects with risk factor. *Represents variables with unadjusted RR more than 1.5 that were selected for entry into the stepwise forward and backward proportional hazard model. †Results of age-adjusted stepwise proportional hazards model.

cognitive performance, the incidence of delirium was significantly increased in those with vascular risk factors (17% vs 6%, P ⬍ .01). In subjects without vascular risk factors, the incidence of delirium was increased in subjects with mildly impaired cognitive performance compared with those without (17% vs 6%, P ⬍ .01). The incidence of delirium was highest in subjects with tobacco exposure who were undergoing vascular surgery and who also had mildly impaired cognitive performance (38%, P ⫽ .03). This suggests a “double gradient effect” in which both vascular and cognitive performance factors contribute to delirium in independent and important ways.29 The independent contributions of vascular surgery, mildly impaired cognitive performance, and age to delirium risk are presented in Table 5. The presence of both independent vascular risk factors tripled the risk of delirium

Table 4

(adjusted RR 3.2, 95% CI, 2.1-4.9) independently of age and mildly impaired cognitive performance. Mildly impaired cognitive performance more than doubled the risk of delirium (adjusted RR 2.2, 95% CI, 1.4-3.6). Advancing age also was significantly associated with the risk of delirium (adjusted RR 1.1, 95% CI, 1.0-1.1 per year).

DISCUSSION In this study assessing preoperative risk factors for delirium in patients undergoing noncardiac surgery, tobacco exposure, the need for vascular surgery, and mildly impaired cognitive performance on 2 measures of executive function contributed independently to delirium risk. Many studies have identified individual vascular risk factors for delirium,1,3,9,10 but our study is unique in exam-

Relationship of Vascular and Cognitive Risk for Delirium Mildly Impaired Cognitive Performance†

Vascular risk factors*

0 or 1 n/N (%) 2 n/N (%)

No n/N (%)

Yes n/N (%)

50/903 (5.5%) 23/133 (17.3%)

16/93 (17.2%) 10/26 (38.5%)

n ⫽ number of subjects with delirium; N ⫽ number of subjects with risk factor. *Independent vascular risk factors include tobacco exposure and vascular surgery. Subjects with no vascular risk factor were as likely to develop delirium as those with 1 vascular risk factor (5.4% vs 7.6%, P ⫽ .18). †Mildly impaired cognitive performance is defined as performance ⱕ 1.5 SD below the mean on either of the 2 independent tests associated with delirium (list them here). Subjects with borderline performance (ⱕ1.5 SD below the mean) on 1 of the tests were as likely to develop delirium as those with borderline performance on both tests (21% vs 33%, P ⫽ .31).

812

The American Journal of Medicine, Vol 120, No 9, September 2007

Table 5 Independent Relationship of Vascular and Cognitive Risk Factors for Delirium Variable

Adjusted RR (95% CI)

Vascular risk factors* Mildly impaired cognitive performance† Age (per y)

3.2 (2.1-4.9) 2.2 (1.4-2.7) 1.1 (1.0-1.1)

RR ⫽ relative risk; CI ⫽ confidence interval. *Describes the risk for delirium in subjects with both independent vascular risk factors (tobacco exposure and vascular surgery) to those with none or 1 vascular risk factor. †Describes the risk for delirium in subjects with cognitive performance ⱕ 1.5 SD below the mean on both or either of the independent neuropsychologic tests to those with cognitive performance ⬎ 1.5 SD below the mean.

ining multiple vascular risk factors in a broad spectrum of noncardiac surgery patients. We identified tobacco exposure and vascular surgery to be independent risk factors for delirium. Our findings are consistent with previous studies that have identified tobacco use10 and vascular surgery1 as risk factors for delirium. Although no study has determined whether the increased delirium risk is attributable to atherosclerosis predisposition or the surgical procedure itself,11,30 previous evidence and our study results support increased atherosclerosis burden as a risk factor for postoperative delirium.9 Atherosclerosis is the most common disease process affecting older adults and is not isolated to arteries of the heart, but also affects arteries (and physiologic function) in the brain, kidneys, and gastrointestinal tract. Increasing evidence suggests that atherosclerosis risk factors31,32 and pathology33 affect frontal lobe control processes called executive functions,25,34,35 although other pathologies such as Alzheimer disease have been implicated.36 Executive function is a broad term for functions such as planning, working memory, impulse control, inhibition, and sequencing, as well as the selective attention that is closely related, but not exclusively localized, to the frontal lobe.25 Recent studies have found that executive function impairment is associated with increased cerebral vascular disease in the white matter connections to the frontal lobes (leukoaraiosis).35 The diagnosis of delirium is defined by a deficit in attention,24 which also is an executive function. Recent work has identified poor executive function performance as a risk for delirium after cardiac surgery.15 Because of the redundancy and complexity of the frontal lobe, early-stage executive function changes often go unnoticed unless directly assessed by a neuropsychologic battery.25 By using neuropsychologic assessments to examine multiple cognitive domains, we identified 2 neuropsychologic measures of executive function25 that were independently associated with delirium. This finding may refine the size and scope of neuropsychologic batteries used to further define the relationship of mildly impaired cognitive performance and delirium after surgery. In addition, cognitive

impairment caused by atherosclerosis is highly prevalent and is a spectrum phenomenon, with mild impairment more common than dementia.13 Future work will build on our findings to determine how the level of cognitive impairment increases delirium risk. This study has several strengths and limitations that deserve comment. The large, prospectively enrolled sample from Europe and the United States included many noncardiac surgical procedures. However, data on ethnicity were not collected, which may limit generalizability. Subjects with dementia, cognitive impairment at baseline (MMSE ⬍ 24), or prior neuropsychologic testing were excluded, and thus, our findings are applicable to those without signs of gross cognitive impairment. Overall, baseline neuropsychologic assessment was performed on most subjects (98% completion), and delirium follow-up was completed on 96%. The study began enrollment more than 10 years ago, and surgical techniques, postoperative management, and methods for the assessment and diagnosis of delirium have improved. The criteria for the diagnosis of delirium in the ISPOCD are based on the DSM-III.24 Current criteria for delirium diagnosis are based on DSM-IIIr or DSM-IV and have been demonstrated to be more sensitive for the diagnosis of delirium.37 The diagnostic criteria have increased sensitivity and specificity when completed after completion of a mental status interview, an attention assessment, a delirium severity rating scale, and a standardized delirium instrument.3,38,39 The ISPOCD assessment used the orientation items of the MMSE without an attention assessment or a delirium instrument. Thus, the delirium rate in our study likely represents an underestimate of the true rate.

CONCLUSIONS Delirium has been well recognized to pose a major risk in postoperative patients and is associated with substantial morbidity and mortality. The identification of vascular risk factors and mildly impaired cognitive performance as independent risk factors for delirium holds substantial clinical implications. Use of these risk factors will help to identify patients at high risk for postoperative delirium, who may benefit from interventions to prevent delirium and its attendant complications. Moreover, future trials targeted specifically toward vascular risk factors and improving executive dysfunction may help to decrease delirium in high-risk populations and, ultimately, to improve quality of life for these patients and their families.

ACKNOWLEDGMENTS We thank the ISPOCD group for access to their data.

References 1. Marcantonio ER, Goldman L, Mangione CM, et al. A clinical prediction rule for delirium after elective noncardiac surgery. JAMA. 1994; 271:134-139.

Rudolph et al

Vascular Risk Factors for Delirium

2. Yildizeli B, Ozyurtkan MO, Batirel HF, et al. Factors associated with postoperative delirium after thoracic surgery. Ann Thorac Surg. 2005; 79:1004-1009. 3. Inouye SK. Delirium in older persons. N Engl J Med. 2006;354:11571165. 4. Franco K, Litaker D, Locala J, Bronson D. The cost of delirium in the surgical patient. Psychosomatics. 2001;42:68-73. 5. Marcantonio ER, Flacker JM, Michaels M, Resnick NM. Delirium is independently associated with poor functional recovery after hip fracture. J Am Geriatr Soc. 2000;48:618-624. 6. McCusker J, Cole M, Abrahamowicz M, et al. Delirium predicts 12-month mortality. Arch Intern Med. 2002;162:457-463. 7. Levkoff SE, Evans DA, Liptzin B, et al. Delirium. The occurrence and persistence of symptoms among elderly hospitalized patients. Arch Intern Med. 1992;152:334-340. 8. Rudolph JL, Marcantonio ER. Caring for the patient with postoperative delirium. Hospitalist. 2004;8:20-25. 9. Rudolph JL, Babikian VL, Birjiniuk V, et al. Atherosclerosis is associated with delirium after coronary artery bypass graft surgery. J Am Geriatr Soc. 2005;53:462-466. 10. Dubois MJ, Bergeron N, Dumont M, et al. Delirium in an intensive care unit: a study of risk factors. Intensive Care Med. 2001;27:12971304. 11. Benoit AG, Campbell BI, Tanner JR, et al. Risk factors and prevalence of perioperative cognitive dysfunction in abdominal aneurysm patients. J Vasc Surg. 2005;42:884-890. 12. Fick DM, Agostini JV, Inouye SK. Delirium superimposed on dementia: a systematic review. J Am Geriatr Soc. 2002;50:1723-1732. 13. Petersen RC. Mild cognitive impairment as a diagnostic entity. J Intern Med. 2004;256:183-194. 14. Dasgupta M, Dumbrell AC. Preoperative risk assessment for delirium after noncardiac surgery: a systematic review. J Am Geriatr Soc. 2006;54:1578-1589. 15. Rudolph JL, Jones RN, Grande LJ, et al. Impaired executive function is associated with delirium after coronary artery bypass graft surgery. J Am Geriatr Soc. 2006;54:937-941. 16. Moller JT, Cluitmans P, Rasmussen LS, et al. Long-term postoperative cognitive dysfunction in the elderly ISPOCD1 study. ISPOCD investigators. International Study of Post-Operative Cognitive Dysfunction. Lancet. 1998;351:857-861. 17. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state.” A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12:189-198. 18. Rasmussen LS, Larsen K, Houx P, et al. The assessment of postoperative cognitive function. Acta Anaesthesiol Scand. 2001;45:275-289. 19. Brand N, Jolles J. Learning and retrieval rate of words presented auditorily and visually. J Gen Psychol. 1985;112:201-210. 20. Houx PJ, Vreeling FW, Jolles J. Rigorous health screening reduces age effect on memory scanning task. Brain Cogn. 1991;15:246-260. 21. Wechsler D. Wechsler Adult Intelligence Scale-Revised Manual. New York: Psychological Corporation, A Harcourt Assessment Company; 1989.

813 22. Houx PJ, Jolles J. Age-related decline of psychomotor speed: effects of age, brain health, sex, and education. Percept Mot Skills. 1993;76:195211. 23. Cattell RB, Cattell AKS. The Individual or Group Culture Fair Intelligence Test. Institute for Personality and Ability Testing. Champaign, IL: University of Illinois; 1960. 24. Diagnostic and Statistical Manual of Mental Disorders. 3rd edition. Washington, DC: American Psychiatric Association; 1980. 25. Stuss DT, Levine B. Adult clinical neuropsychology: lessons from studies of the frontal lobes. Annu Rev Psychol. 2002;53:401-433. 26. D’Agostino RB, Sr, Grundy S, Sullivan LM, Wilson P. Validation of the Framingham coronary heart disease prediction scores: results of a multiple ethnic groups investigation. JAMA 2001;286:180-187. 27. Prescott E, Hippe M, Schnohr P, et al. Smoking and risk of myocardial infarction in women and men: longitudinal population study. BMJ 1998;316:1043-1047. 28. Camargo CA Jr, Stampfer MJ, Glynn RJ, et al. Prospective study of moderate alcohol consumption and risk of peripheral arterial disease in US male physicians. Circulation. 1997;95:577-580. 29. Feinstein AR. Clinimetrics. New Haven, CT: Yale University Press; 1987. 30. Bohner H, Hummel TC, Habel U, et al. Predicting delirium after vascular surgery: a model based on pre- and intraoperative data. Ann Surg. 2003;238:149-156. 31. Pugh KG, Kiely DK, Milberg WP, Lipsitz LA. Selective impairment of frontal-executive cognitive function in African Americans with cardiovascular risk factors. J Am Geriatr Soc 2003;51:1439-1444. 32. Kuo H, Iloputaife I, Gagnon M, et al. Effect of blood pressure on cognitive functions in the elderly. J Am Geriatr Soc 2004;59:11911194. 33. O’Sullivan M, Morris RG, Huckstep B, et al. Diffusion tensor MRI correlates with executive dysfunction in patients with ischaemic leukoaraiosis. J Neurol Neurosurg Psychiatry. 2004;75:441-447. 34. Kuo HK, Lipsitz LA. Cerebral white matter changes and geriatric syndromes: is there a link? J Gerontol A Biol Sci Med Sci. 2004;59: 818-826. 35. Price CC, Jefferson AL, Merino JG, et al. Subcortical vascular dementia: integrating neuropsychological and neuroradiologic data. Neurology. 2005;65:376-382. 36. Johnson JK, Vogt BA, Kim R, et al. Isolated executive impairment and associated frontal neuropathology. Dement Geriatr Cogn Disord. 2004;17:360-367. 37. Laurila JV, Pitkala KH, Strandberg TE, Tilvis RS. Delirium among patients with and without dementia: does the diagnosis according to the DSM-IV differ from the previous classifications? Int J Geriatr Psychiatry. 2004;19:271-277. 38. Inouye SK, Foreman MD, Mion LC, Katz KH, et al. Nurses’ recognition of delirium and its symptoms: comparison of nurse and researcher ratings. Arch Intern Med. 2001;161:2467-2473. 39. Simon SE, Bergmann MA, Jones RN, et al. Reliability of a structured assessment for nonclinicians to detect delirium among new admissions to postacute care. J Am Med Dir Assoc. 2006;7:412-415.

The American Journal of Medicine (2007) 120, 814-818

CLINICAL RESEARCH STUDY

Criteria for the Electrocardiographic Diagnosis of Atrial Flutter Improve Diagnostic Accuracy Kenneth M. Weinberg, MD, Pablo Denes, MD, Alan H. Kadish, MD, Jeffrey J. Goldberger, MD Department of Medicine and the Feinberg Cardiovascular Institute, Northwestern University, Chicago, Ill. ABSTRACT PURPOSE: We previously developed and validated diagnostic criteria for the differentiation of atrial flutter from atrial fibrillation. In this study we examine if the criteria (F waves in the frontal plane and a partially or completely regular ventricular response) can improve the diagnostic accuracy of internists. METHODS: Two groups of 10 internists (1 group given the criteria and 1 not) read a set of electrocardiograms (ECGs) selected from the hospital database with cardiologist-confirmed diagnoses of atrial fibrillation, atrial flutter, or “atrial fibrillation-flutter” (100 each). The final diagnoses of all ECGs were provided by a consensus of electrophysiologists. The criteria also were used to establish the criteria-based diagnoses. RESULTS: Of the 298 ECGs analyzed, the electrophysiologist diagnosis was atrial fibrillation in 71% and atrial flutter in 29%. The concordance of the internists’ diagnoses with the electrophysiologist consensus diagnoses was 66 ⫾ 12% for those not given the criteria and 81 ⫾ 4% (P ⬍.01) for those given the criteria. The concordance of the internists’ diagnoses with the criteria based diagnoses was 66 ⫾ 12% for those not given the criteria and 83 ⫾ 4% (P ⬍.01) for those given the criteria. CONCLUSIONS: The simple criteria of F waves in the frontal plane and a partially or completely regular ventricular response can be used to improve the differentiation of atrial flutter from atrial fibrillation based on the ECG. © 2007 Elsevier Inc. All rights reserved. KEYWORDS: Atrial flutter; Atrial fibrillation; Electrocardiogram

Atrial flutter is the second most common atrial tachyarrhythmia after atrial fibrillation, with an estimated 200,000 new cases in the United States annually.1 Heretofore, no validated criteria have been identified to accurately make this diagnosis. Typically, the diagnosis is made by the identification of a “saw-tooth” pattern.2 However, this imprecise criterion frequently leads to incorrect diagnoses. We3 and others4 have shown that significant disagreement exists even among experienced ECG readers. For instance, we have demonstrated a diagnostic concordance of only 85%, even among cardiologists.3 ECGs with some features of atrial flutter and Requests for reprints should be addressed to Jeffrey J. Goldberger, MD, Northwestern Memorial Hospital, 251 East Huron Street Feinberg 8-542, Chicago, IL 60611. E-mail address: [email protected]

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2006.06.013

some features of atrial fibrillation are the primary source of disagreement. To improve diagnostic accuracy, we sought to develop and validate criteria3 for the diagnosis of atrial flutter. The simple criteria of F waves in the frontal plane and a partially or completely regular ventricular response had a sensitivity of 90% and specificity of 100% based on arrhythmia diagnoses confirmed with intracardiac recordings. These criteria were tested in a second set of unselected ECGs with diagnoses of atrial fibrillation and atrial flutter; the criteria had an overall concordance with the consensus electrophysiologist diagnoses of 94%. Internists are often the first health care providers to whom patients with atrial arrhythmia present. The ability to correctly diagnose these tachycardias may have a significant impact on the type of care subsequently provided. There-

Weinberg et al

ECG Diagnosis of Atrial Flutter

fore, in this study, we further tested the applicability and utility of these criteria among internists.

METHODS

815 partially or completely regular ventricular response. A set of instructions is shown in Figure 1. No example ECGs were given. The other group of 10 internists were not given any instructions and were simply asked to provide a rhythm diagnosis.

This study was approved by the Northwestern University InstituStatistical Analysis tional Review Board. Three hunFor continuous variables, mean CLINICAL SIGNIFICANCE dred ECGs were obtained from and standard deviations were calthe Marquette Muse database at culated and expressed in appropri● Errors in differentiating atrial flutter Northwestern Memorial Hospital. ate units. Student’s t test was used from coarse atrial fibrillation are not unA preliminary interpretation of to compare means for continuous common and can lead to alterations in these ECGs, including a rhythm variables. A P value ⬍.05 was management. diagnosis, is supplied by the comconsidered significant. puter; the ECGs are then overread ● Using the criteria of F waves in the fronby cardiologists who assign a final tal plane and a partially or completely diagnosis, which is stored in the RESULTS regular ventricular response can improve database. ECGs were selected so Of the 300 ECGs, 2 were removed diagnostic accuracy for atrial flutter. as to find 100 each of the cardiolfrom further analysis due to techogist-confirmed diagnoses of nically poor tracings. The subjects atrial fibrillation and atrial flutter. whose ECGs were included had Over the period reviewed from the database, 6.4% of ECGs an average age of 68 ⫾ 14 years and 60% were male. Table were diagnosed as atrial fibrillation and 1% were diagnosed 1 provides the breakdown of diagnoses based on both the as atrial flutter. In order to identify a set of ECGs in which criteria and the electrophysiologist consensus diagnosis for the diagnosis of atrial fibrillation or atrial flutter is particuthe ECGs included in this study, which had already been larly difficult to make, an additional 100 ECGs with the interpreted by an individual cardiologist. Cardiologists and cardiologist-confirmed diagnosis of “atrial fibrillation-flutboth the electrophysiologist consensus and criteria-based ter” were identified. Although textbooks usually only briefly diagnoses had a high concordance for the diagnosis of atrial mention atrial fibrillation-flutter, also called impure flutter, fibrillation, but the cardiologists tended to use the diagnosis it is sometimes used as a term to describe ECGs with of atrial flutter more liberally than the electrophysiologists relatively large amplitude fibrillatory waves. Wagner2 refers or the criteria. Specifically, of the 100 ECGs interpreted by to a “spectrum” of atrial fibrillation—atrial flutter-fibrillathe cardiologists as atrial flutter, the electrophysiologists’ tion—atrial flutter. Although it is possible to have atrial consensus and criteria-based diagnoses were atrial fibrillaflutter in one atrium and atrial fibrillation in the other, this tion rather than atrial flutter in 30 and 37, respectively. Of is likely a rare event5 and even when it occurs, the ECG note, the vast majority (86%) of ECGs interpreted as “atrial fibrillation-flutter” by the cardiologists were diagnosed as pattern typically follows the rhythm abnormality noted in atrial fibrillation by both electrophysiologists and the the right atrium. Although the ECG diagnosis of atrial criteria. fibrillation-flutter has no clear pathophysiologic correlate, it is a diagnosis often used to describe ECGs with some Concordance of Internists’ Diagnoses with features of atrial fibrillation and some features of atrial flutter. Over the period reviewed to obtain these ECGs, Criteria-Based Diagnoses 0.08% of the total ECGs were classified as atrial fibrillationThe internists’ concordance with the criteria-based diagflutter. Because the presence of paced rhythms or premature noses was 66% ⫾ 12% for those not given the criteria and ventricular complexes might affect the ability to classify the 83% ⫾ 4% (P ⬍.01) for those given the criteria. regularity of the ventricular response, ECGs with these The criteria classified 79 ECGs as atrial flutter. Mean age attributes were excluded from analysis. Consequently, 364 was 66 ⫾ 11 years, and 77% were male. For these ECGs, the ECGs were screened to find 300 acceptable ECGs. concordance of the internists not given the criteria was The ECGs were read blindly (blind to the cardiologist’s 80 ⫾ 11% and concordance of the internists given the cridiagnosis) by 2 cardiac electrophysiologists. In cases of teria was 76 ⫾ 9%, which was not significantly different. disagreement (42/300 cases) between the 2 readers, a third The criteria classified 219 ECGs as atrial fibrillation. electrophysiologist read the ECG, and a consensus diagnoMean age was 69 ⫾ 15 years, and 53% were male. For these sis was agreed to by all. Each ECG also was assigned a ECGs, the concordance of the internists not given the cridiagnosis based on the criteria. In order to evaluate the teria was 61% ⫾ 19%, and concordance of the internists criteria, 2 groups of 10 internists each blindly read all the given the criteria was 85% ⫾ 8% (P ⬍.01). ECGs. One group of 10 internists was given instructions Figure 2 shows an example of an ECG that was read by with the 2 criteria for the diagnosis of atrial flutter identified the cardiologist as atrial flutter. Seven of 10 internists not given the criteria also interpreted it as atrial flutter. The in our validation study: F waves in the frontal plane and a

816

The American Journal of Medicine, Vol 120, No 9, September 2007

Figure 1

Instructions for applying the criteria as given to one set of internists.

criteria-based diagnosis was atrial fibrillation, as was the electrophysiologist consensus diagnosis. All 10 internists given the criteria diagnosed atrial fibrillation.

Concordance of Internists’ Diagnoses with Electrophysiologist Consensus Diagnoses The internists’ concordance with the electrophysiologists diagnoses was 66% ⫾ 12% for those not given the criteria and 81% ⫾ 4% (P ⬍.01) for those given the criteria. The electrophysiologists classified 85 ECGs as atrial flutter. Mean age was 66 ⫾ 13 years, and 74% were male. Of these ECGs, the criteria diagnosed atrial flutter in 73, for a concordance of 86%. For these ECGs, the concordance of the internists not given the criteria was 78% ⫾ 10% and concordance of the internists given the criteria was 71% ⫾ 4%, which was not significantly different (P ⫽ NS). The electrophysiologists classified 213 ECGs as atrial fibrillation. Mean age was 69 ⫾ 15 years, and 54% were male. Of these ECGs, the criteria diagnosed atrial fibrillation in 207 for a concordance of 97%. For these ECGs, the concordance of the internists not given the criteria was 62% ⫾ 20% and concordance of the internists given the criteria was 85% ⫾ 7% (P ⬍.01).

Table 1 Groups

Comparison of ECG Diagnoses among Reading

Cardiologist Diagnosis

Electrophysiologist Consensus Diagnosis

CriteriaBased Diagnosis

AF

AFL

AF

AFL

1 70 14

97 37 85

2 63 14

Atrial fibrillation (n ⫽ 99) 98 Atrial flutter (n ⫽ 100) 30 Atrial fibrillation-flutter 85 (n ⫽ 99)

DISCUSSION The main findings of the present study are that the criteria of F waves in the frontal plane and a partially or completely regular ventricular response provide a simple set of guidelines that aid ECG interpretation and improve internists’ ability to differentiate atrial flutter from atrial fibrillation. The criteria improve diagnostic accuracy predominantly by establishing the 2 criteria necessary for a diagnosis of atrial flutter. In the absence of these criteria, both internists and cardiologists tend to diagnose atrial flutter correctly but tend to classify some cases of atrial fibrillation as atrial flutter. This occurs when there are characteristics of both arrhythmias on an ECG. The criteria can be applied by internists to improve their diagnostic accuracy, particularly in difficultto-interpret ECGs such as those that were labeled atrial fibrillation-flutter by the cardiologists. We propose that these 2 simple criteria be adopted as diagnostic aids. The lack of specific and simple diagnostic criteria in ECG interpretation often leads to inaccurate interpretation. The accuracy of diagnostic ECG interpretations by various groups of physicians has been studied previously, most often in the setting of suspected acute myocardial infarctions in emergency departments.6 Several studies7-10 evaluated overall ECG interpretation in nonselected emergency department patients. In these studies, there was an overall agreement with the reference standard (usually a cardiologist) in only 42%-85% of cases. Diagnostic errors led to changes in management in 0.5%-3.2% of cases. Several previous reports have evaluated whether formal instruction in ECG interpretation can improve diagnostic accuracy. White et al8 showed that formal instruction by a cardiologist decreased the rate of serious errors from 21.3% to 10.7%. In a study by Gjorup et al,11 one group of physicians underwent an 8-hour training program in the interpretation of ECGs for acute myocardial infarction. The experimental group did not improve their diagnostic reading skills, and the authors postulated that the physicians’ diag-

Weinberg et al

ECG Diagnosis of Atrial Flutter

817

Figure 2 An ECG originally read as atrial flutter by a cardiologist. Both the electrophysiologist consensus diagnosis and the criteria-based diagnosis were atrial fibrillation. Internists not given the criteria mostly provided a diagnosis of atrial flutter (7/10) but internists with the criteria all diagnosed the ECG as atrial fibrillation.

nostic practices may be formulated at a very early stage in training. The physicians’ own training and experience may have led them to override the information provided in the training. The present study provides some support for this hypothesis. Despite instructions to use the criteria, the internists had a concordance with the criteria of only 83%. Thus, their ingrained training still caused them to rely on their visual assessment rather than the supplied criteria. The variable appearance of both atrial flutter and atrial fibrillation on the surface ECG makes defining diagnostic criteria difficult. Textbooks of electrocardiography2,12-14 all have similar themes in differentiating one arrhythmia from another, but usually use imprecise terms while pointing out that there can be variants. Although the classic “saw-tooth” pattern of atrial flutter can be described relatively clearly, the authors are quick to state that certain ECGs may not be amenable to easy interpretation due to the QRS or T wave obscuring atrial activity, unusually small flutter waves, or large fibrillatory waves. Rankin and Rae12 describe the saw-tooth flutter wave with a regular rate ⬎250 beats/min. Chung13 describes a regular asymmetric saw-tooth wave with rates ranging from ⬎200 to ⬍350 beats/min. Chou14 takes the appearance of the flutter wave one step further by describing a gradual downward slope followed by an abrupt upward inscription. Fibrillatory waves of atrial fibrillation are also variably defined and have even been classified into fine or coarse depending on their amplitude. However, Chou’s14 definition is most complete: varying amplitude, duration, and morphology that cause a random oscillation of the baseline with rates between 400 and 700 beats/min. These descriptions provide a reader clues to a visual analysis of the ECG but do not provide specific criteria. As the visual discrimination between atrial flutter and atrial fibrillation frequently results in discrepancies even among experts, we propose that these criteria should be adopted as diagnostic aids that have high sensitivity and specificity in the differentiation of these rhythms. As shown by the group

of internists who were given instructions, these criteria can be easily applied, resulting in enhanced ECG reading accuracy. As the treatment approaches for atrial flutter may differ from that for atrial fibrillation, it is important to adopt these new ECG diagnostic criteria.

LIMITATIONS As noted, even with criteria, the concordance of the internists with the electrophysiologist diagnoses was only 81%. This was in large part due to the 83% concordance with the criteria-based diagnoses. Internists given the criteria frequently overrode the diagnosis suggested by the criteria to insert their instinctive visual determination. With dissemination of the validity and utility of the criteria, there should be better acceptance of the criteria-based diagnosis, which should result in substantially improved diagnostic accuracy. It should be noted that there are multiple types of atrial flutter, with the main distinction being between those that are dependent on the cavotricuspid isthmus (isthmus-dependent, also referred to as common or typical atrial flutter) and those that are not isthmus-dependent (atypical atrial flutter). In our validation study,3 the vast majority of atrial flutters with intracardiac electrophysiologic confirmation of the diagnosis were isthmus-dependent. However, we also studied a large unselected set of ECGs with atrial flutter, in which intracardiac mapping studies were not performed. Although the exact distribution of these types of flutter is not known, it is likely that a significant proportion were not isthmusdependent. In any case, the presence of a single macroreentrant circuit in the atrium, regardless of its dependence on the cavotricuspid isthmus, should conform to the proposed criteria, whereas atrial fibrillation does not. However, it is important to note that the diagnostic criteria do not have 100% sensitivity and specificity. Thus, experienced ECG readers may still need to override the criteria-based diag-

818

The American Journal of Medicine, Vol 120, No 9, September 2007

noses when there are other visual clues on the ECG suggesting an alternative diagnosis.

IMPLICATIONS Differentiating atrial flutter from atrial fibrillation is more important than ever today as treatment strategies for these arrhythmias may differ. Atrial flutter, as a macroreentrant circuit located in the right atrium, can be cured with a high chance of success and a low complication rate by radiofrequency catheter ablation,15,16 which is often chosen as first line therapy, particularly because this appears to be preferable to medical therapy.17 Although catheter-based therapeutic techniques for atrial fibrillation are gaining in popularity, especially for paroxysmal forms caused by ectopic tachycardias originating from the pulmonary veins, their success and complication rates are not nearly as favorable as those for atrial flutter.18 The recent publication of the AFFIRM19 and RACE20 trials may lead to increased use of the rate control strategy in patients with atrial fibrillation. Although there are no studies comparing catheter ablation for atrial flutter with a rate control strategy, given the high success rate of catheter ablation and its low morbidity, it seems likely that this could be a superior approach to rate control in these patients. Therefore, correctly differentiating these rhythms can have important therapeutic consequences. The simple criteria of F waves in the frontal plane and a partially or completely regular ventricular response can be used by internists to improve their differentiation of atrial flutter from atrial fibrillation. It is clear that the proposed criteria provide an important new diagnostic aid to differentiate atrial flutter and atrial fibrillation.

References 1. Granada J, Uribe W, Chyou PH, et al. Incidence and predictors of atrial flutter in the general population. J Am Coll Cardiol. 2000;36:22422246. 2. Wagner G. Marriott’s Practical Electrocardiography, 10th ed. Philadelphia, PA: Lippincott Williams & Williams; 2001. 3. Weinberg K, Denes P, Kadish A, Goldberger J. Development and validation of diagnostic criteria for atrial flutter on the surface electrocardiogram. Ann Noninvasive Electrophysiol. In press. 4. Knight BP, Michaud GF, Strickberger SA, Morady F. Electrocardiographic differentiation of atrial flutter from atrial fibrillation by physicians. J Electrocardiol. 1999;32:315-319.

5. Horvath G, Goldberger JJ, Kadish AH. Simultaneous occurrence of atrial fibrillation and atrial flutter. J Cardiovasc Electrophysiol. 2000; 11:849-858. 6. Salerno SM, Alguire PC, Waxman HS. Competency in interpretation of 12-lead electrocardiograms: a summary and appraisal of published evidence. Ann Intern Med. 2003;138:751-760. 7. Todd KH, Hoffman JR, Morgan MT. Effect of cardiologist ECG review on emergency department practice. Ann Emerg Med. 1996;27: 16-21. 8. White T, Woodmansey P, Ferguson DG, Channer KS. Improving the interpretation of electrocardiographs in an accident and emergency department. Postgrad Med J. 1995;71:132-135. 9. Westdrop EJ, Gratton MC, Watson WA. Emergency department interpretation of electrocardiograms. Ann Emerg Med. 1992;21: 541-544. 10. Kuhn M, Morgan MT, Hoffman JR. Quality assurance in the emergency department: evaluation of the ECG review process. Ann Emerg Med. 1992;21:10-15. 11. Gjorup T, Kelbaek H, Nielsen D, Kreiner S, Godtfredsen J. Interpretation of the electrocardiogram in suspected myocardial infarction: a randomized controlled study of the effect of a training programme to reduce interobserver variation. J Intern Med. 1992;231:407-412. 12. Rankin A, Rae A. Sinus and atrial arrhythmias. In: MacFarlane P, Veitch Lawrie TD, eds. Comprehensive Electrocardiology: Theory and Practice in Health and Disease, vol. 2. New York, NY: Pergamon Press; 1989:881-914. 13. Chung E. Electrocardiography: Practical Applications with Vectorial Principles. New York, NY: Appleton-Century-Crofts; 1985. 14. Chou T, Knilans T. Electrocardiography in Clinical Practice: Adult and Pediatric. Philadelphia, PA: W.B. Saunders; 1996. 15. Poty H, Saoudi N, Nair M, Anselme F, Letac B. Radiofrequency catheter ablation of atrial flutter. Further insights into the various types of isthmus block: application to ablation during sinus rhythm. Circulation. 1996;94:3204-3213. 16. Tai CT, Chen SA, Chiang CE, et al. Long-term outcome of radiofrequency catheter ablation for typical atrial flutter: risk prediction of recurrent arrhythmias. J Cardiovasc Electrophysiol. 1998;9:115-121. 17. Natale A, Newby KH, Pisano E, et al. Prospective randomized comparison of antiarrhythmic therapy versus first-line radiofrequency ablation in patients with atrial flutter. J Am Coll Cardiol. 2000;35:18981904. 18. Padanilam BJ, Prystowsky EN. Should atrial fibrillation ablation be considered first-line therapy for some patients? Should ablation be first-line therapy and for whom? the antagonist position. Circulation. 2005;112:1223-1229; discussion 1230. 19. Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) Investigators. A comparison of rate control and rhythm control in patients with atrial fibrillation. N Engl J Med. 2002;347: 1825-1833. 20. Van Gelder IC, Hagens VE, Bosker HA, et al. A comparison of rate control and rhythm control in patients with recurrent persistent atrial fibrillation. N Engl J Med. 2002;347:18341840.

The American Journal of Medicine (2007) 120, 819.e9-819.e11

BRIEF OBSERVATION

Medical Errors Arising from Outsourcing Laboratory and Radiology Services Brian S. Chasin, MD,a Sean P. Elliott, MD,b Stephen A. Klotz, MDa Sections of Infectious Diseases, aDepartment of Medicine and bDepartment of Pediatrics, University of Arizona, Tucson. ABSTRACT OBJECTIVE: Document errors and the nuisance factor inherent in the informational exchange that occurs with the outsourcing of laboratory and radiology examinations. METHODS: Three infectious diseases physicians at a tertiary care hospital recorded problems involving data transmitted by telephone or fax from outsource providers for 4 months. This included in- and outpatients, and those in transition from one status to another. RESULTS: Outsourcing laboratory and radiology examinations of insured outpatients is a common practice. Insurance companies determine which healthcare facility performs these tests based on contractual agreements with outsource providers. This leads to confusion and frustration for the doctor and patient alike, and occasionally, to medical error. CONCLUSIONS: The exchange of patient data involved in outsourcing is subject to systemic errors that do not allow of easy solution. © 2007 Elsevier Inc. All rights reserved. KEYWORDS: Outsourcing laboratory; Outsourcing radiology; Medical error

Insurance companies regularly subcontract with other corporate entities for services such as radiology and laboratory examinations for outpatients. Consequently, insurance companies determine “when and where” diagnostic tests are performed. Insured outpatients seen in large hospitals equipped with the technology to perform requested tests are frequently required to travel to other locations to have the test performed. This is required for the patient to qualify for insurance coverage of the requested test. For example, Insurance Company A allows a hospital-based physician to obtain laboratory tests on an outpatient in the hospital clinic, whereas Insurance Company B, which does not have a contract with the hospital for outpatient laboratory examinations, requires the outpatient to travel to another location to obtain the test. These rules are unyielding and may lead to incongruous situations. For example, a patient hospitalized for weeks is discharged to be followed in the hospital’s outpatient clinic. However, due to contractual agreements, Requests for reprints should be addressed to Stephen A. Klotz, MD, Department of Medicine, University of Arizona, 1501 N. Campbell Ave. Tucson, AZ 85724. E-mail address: [email protected]

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2006.07.024

the patient, now an outpatient, is ineligible for laboratory or radiology services performed in the hospital and must travel to another site to have the tests performed. Off-site laboratory and radiology services perform requested tests in a timely fashion. However, where requested examinations are performed is fraught with difficulties. We describe situations illustrating the frustrations and mistakes involved in outsourcing laboratory and radiology services.

METHODS There are 10 hospitals in Tucson, Arizona with full-service radiology and laboratory services, and most physicians associated with these hospitals conduct outpatient clinics in the hospitals or in adjacent buildings. The preponderance of outsourced radiology tests go to Radiology Ltd. (Tucson), which has 10 freestanding sites, and laboratory tests to Sonora Quest Laboratories of Tucson (Tucson), with 8 sites, or Laboratory Corporation of America (Burlington, NC), with 11 sites. The study was conducted at University Medical Center, the major teaching hospital for the University of Arizona Health Sciences Center. It has 365 beds and more than 450 credentialed physicians, most of whom have outpatient clinics

819.e10

The American Journal of Medicine, Vol 120, No 9, September 2007

in the hospital proper. The clinical laboratory and radiology services are located in the hospital and are fully accredited with their respective accrediting organizations. Three physicians documented incidents involving outpatient laboratory and radiology requests over a 4-month period of time.

Microbiology Resource Committee of the College of American Pathologists recommends that all but the smallest hospital laboratories should not outsource bacterial cultures and commented that “It is worse to provide an incorrect interpretation of a direct smear than to provide no interpretation at all . . .”1

Case 2: An Unnecessary Thoracotomy. A 48-year-old woman, CLINICAL SIGNIFICANCE who was a heavy smoker, was diExamples of Outsourced agnosed with coccidioidomycosis ● This is an original contribution dealing Laboratory and Radiology following biopsy of a pulmonary with the outsourcing of laboratory and Tests Resulting in nodule while an inpatient in hosradiology examinations. It describes Medical Errors pital A. She was placed on fluconerrors that may occur in this system. It Case 1: HIPAA Trumps Cause azole. She was scheduled to reinvolves insurance practices, informatics of Meningitis. A 6-month-old boy ceive follow-up chest radiographs and confidentiality. presented to his primary care phyin the hospital but due to insursician (PCP) with acute onset of ance company policy was required ● Errors in outsourcing may lead to adfever. His clinical appearance was to have these performed at an outverse patient outcomes. consistent with sepsis, and blood side radiology site. Several folfor culture was sent to an off-site ● A more systematic study of this problem low-up chest radiographs were laboratory. The provider adminisis needed. performed, and the official reading tered an intramuscular dose of noted a nodule that was larger ceftriaxone and admitted him to than that described at hospital A. the hospital. A lumbar puncture She was subsequently admitted to revealed cerebrospinal fluid (CSF) with 75 neutrophils, gluhospital B, where the nodule was removed surgically. It cose of 24 mg/dL, and a protein 167 mg/dL. The Gram stain demonstrated only coccidioidomycosis. did not demonstrate bacteria. Vancomycin and ceftriaxone were administered for presumed bacterial meningitis, Comment. This illustrates a common problem. Off-site but vancomycin was discontinued when the off-site laboraradiology examinations for outpatients are dictated by the tory called the PCP and stated that Gram-negative diploradiologists and then faxed to the requesting physician’s cocci were growing from the blood culture, consistent with office. Rarely, if ever, does the PCP even see the radioNeisseria meningitidis. Attempts to confirm the blood culgraph, much less have the ability to compare films from ture results were unavailing as the laboratory refused to different sources. Occasionally, CDs of radiology examinarelease information to anyone but the referring PCP (who tions are sent with outpatients, but more often than not there was not available over the weekend). Forty-eight hours after is no compatible computer program to “read” them. Review admission, the patient developed seizures, respiratory failof the radiographs in this case revealed that the nodule had ure, and shock requiring admission to the Pediatric ICU. not changed in size and that surgery was unnecessary. If the Vancomycin was begun again because of concern that the chest radiograph had been repeated in the same hospital, the blood culture results were in error. The patient stabilized. confusion about the size of the nodule would not have CSF subsequently cultured Gram-positive diplococci, eventuoccurred. ally identified as penicillin- and ceftriaxone-resistant Streptococcus pneumoniae. Six days after admission the initial blood Further Examples of Error Involving Patient Care Due culture results were communicated by letter to the PCP reportto Outsourcing. Two other notable errors occurred in 4 ing S. pneumoniae. The patient survived and has seizures and months. One, a patient with a pituitary abscess, received cognitive delay resulting from his meningitis. prolonged and unnecessary intravenous antibiotics because a repeat magnetic resonance imaging scan could not be Comment. The failure of the off-site laboratory to properly obtained in the same facility for purposes of comparison. identify the causative microorganism, as well as to notify Another patient with coccidioidomycosis requiring hospithe hospital in a timely fashion of the correct identification talization obtained serological tests at 2 laboratory facilities. led directly to inadequate antibiotic coverage for meningitis The results were divergent and therapy was stopped premaand adverse sequelae for this patient. Subsequent review turely, leading to a relapse in disease. revealed that a corrected blood culture result was available at Incomplete or missing reports from outsourcers are comthe commercial laboratory before the CSF culture result obmonplace and potentially place patients at risk (Table 1). tained in the hospital. Had it been released, the patient’s course We studied all laboratory reports faxed from the 2 major might have been altered. The laboratory maintained that Health laboratory outsourcers for HIV-1 infected outpatients and Insurance Portability and Accountability Act (HIPAA) regufound that 47% of the reports were duplicates or pending lations prevented them from providing the information. The results; 9% were reports that contained no relevant patient

RESULTS

Chasin et al Table 1

Outsourcing Laboratory and Radiology Services

819.e11

Examples of Errors Committed at Outsource Laboratories and Their Consequences

Test Ordered

Outsource Laboratory Report

Consequences of the Error and/or Comments

Any blood or urine test

Specimen improperly collected, inadvertently discarded, request overlooked, and results not communicated in timely fashion

Commonplace mistakes; these errors lead to repeating the test, cancellation of a scheduled outpatient visit, and/or delaying or interfering with therapy

Collection date not noted at outsource laboratory site, therefore, culture discarded Test not performed; inappropriate specimen

Lesion disappeared and consequently there was no definitive diagnosis

Specific examples Viral culture for Herpes simplex Brucella serology HIV-1 RNA Nitroblue tetrazolium assay

Wrong test performed; sometimes HIV-1 DNA or HCV RNA are erroneously performed Test not performed; technician unfamiliar with test and unable to identify test code

information, 8% of the reports contained mistakes originating at the outsourcers involving 5 patients (2 specimens were not frozen, one specimen was not collected, one specimen had results that could not be reported because they could not be found, and one patient had the wrong test performed); only 36% of the reports had finalized results.

DISCUSSION Most large clinics have laboratory and radiology facilities on-site with integrated data bases. However, a patient’s insurer may require the outpatient to travel elsewhere for required examinations where the training and skills of the employees are not necessarily equivalent to those of the hospital.2 We describe what occurs when physicians affiliated with hospitals order tests on outpatients who are required by their insurance companies to go off-site for tests. Some of the problems were similar to those previously reported, such as turnaround times for test reports, compromised specimens, incorrect tests, and problems with the courier service.3,4 In addition, we also document significant medical error due to outsourcing of laboratory and radiology requests. It is difficult to resolve problems with outsourcers because their employees are not knowledgeable regarding specific tests, and when the physician finally does find someone who understands the issue, there is little that can be done. This type of problem is exemplified by the last entry in Table 1. Patients are often upset when informed that they must obtain their examinations elsewhere. Not uncommonly, they fail to have the requested test performed in a timely manner, or at all (Table 1, third entry). Physicians need to follow-up

Patient never had the test performed; refused to drive 90 miles to repeat it Physician must recognize the problem and reorder the test Diagnosis of chronic granulomatous disease delayed by 2 months due to inability to perform tests at outsource laboratory (test finally performed elsewhere)

on tests that have been ordered to ensure that they are performed. This requires a sophisticated and complex tracking system.4,5 Addressing the issue of outsourced radiology examinations, Wachter remarked that outsourcing may have virtue but the practice can be harmful if quality is sacrificed.6 Outsourcing has implications beyond monetary savings, and the real financial and human costs must be factored into the overall equation.7 Although some outsourcing is unavoidable, outsourcing to the magnitude practiced in our tertiary care hospital is not likely to be cost-effective given the numerous errors encountered, some of which are grave. A more systematic study of this practice is needed.

References 1. Procop G, Winn W, Microbiology Resource Committee, College of American Pathologists. Outsourcing microbiology and offsite medical education. Implications on patient care, cost savings, and graduate medical education. Arch Pathol Lab Med. 2003;127:623-624. 2. Harty-Golder B. Decentralization skews traditional specimen collection. MLO Med Lab Obs. 2005;37:33. 3. LaBeau K, Granade S, Steindel S. What do you want from your reference laboratory? A report card on the send-out testing in the Pacific Northwest. MLO Med Lab Obs. 2002;33:24-26. 4. Poon EG, Gandhi TK, Sequist TD, Murff HJ, Karson AS, Bates DW. “I wish I had seen this test result earlier!” Dissatisfaction with test result management systems in Primary Care. Arch Int Med. 2004; 164:2223-2228. 5. Elder NC. Patient safety in the physician office setting. AHRQ web M&M, 2006. Available at: http://www.webmm.ahrq.gov/perspective.aspx? perspectiveID⫽24. 6. Wachter R. The “dis-location” of U.S. medicine—the implications of medical outsourcing. N Engl J Med. 2006;354:661-665. 7. Stockamp D. Revenue cycle outsourcing. The real costs and benefits. Healthc Financ Manage 2006;60:84-90.

The American Journal of Medicine (2007) 120, 819.e13-819.e14

BRIEF OBSERVATION

Troponin I Levels in Patients with Preeclampsia Dominique Joyal, MD,a Ferdinand Leya, MD,a Megan Koh, MD,b Richard Besinger, MD,b Ravi Ramana, DO,a Steven Kahn, PhD,c Walter Jeske, PhD,d Bruce Lewis, MD,a Lowell Steen, MD,a Ruben Mestril, PhD,d Dinesh Arab, MDa a

Cardiology Division, bObstetric and Gynecology Division, cPathology Division, and dDivision of Cardiovascular Research, Loyola University Medical Center, Maywood, Ill. ABSTRACT INTRODUCTION: Preeclampsia involves a diffuse inflammatory state and elevated levels of troponins in patients with preeclampsia have been anecdotally reported. It is, however, unknown whether it is attributable to the preeclampsia. OBJECTIVE: We sought to determine the troponin I levels at the time of delivery in pregnant women with and without preeclampsia. METHODS: Plasma samples were obtained at the time of delivery and serum troponin I was measured by ELISA method. RESULTS: Thirty-nine women were included (20 with preeclampsia and 19 without). Mean troponin I level was 0.008 ng/mL in patients with preeclampsia and 0.01 ng/mL in controls (P ⫽.59). The highest troponin I level was 0.04 ng/mL for both patients with and without preeclampsia. CONCLUSIONS: Preeclampsia was not associated with a rise in troponin I levels in our study. Patients with preeclampsia and elevated troponin levels should have further cardiac investigations. © 2007 Elsevier Inc. All rights reserved. KEYWORDS: Troponin; preeclampsia

Cardiac troponins are elevated in several nonatherosclerotic conditions. Besides renal failure, elevated troponins can be seen with tachycardia, pericarditis, heart failure, strenuous exercise, and sepsis.1,2 Preeclampsia is viewed as a systemic disease exhibiting impairment in vascular tone, which can be explained as a maternal response to generalized endothelial dysfunction.3,4 Development of hypertensive disease in pregnancy predicts remote cardiovascular events.5,6 However, it is not clear whether the increased risk of late cardiovascular morbidity reflects an underlying predisposition for both disorders or whether preeclampsia results in permanent arterial changes or myocardial damage. In view of the diffuse inflammatory state induced by preeclampsia, we sought to measure serum troponin levels in patients with and without preeclampsia at the time of delivery.

Requests for reprints should be addressed to Dinesh Arab, MD, Department of Medicine, Division of Cardiology, 2160 South First Avenue, Loyola University Medical Center, Maywood, IL-60153. E-mail address: [email protected]

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2006.05.068

METHODS Between November 2003 and March 2005, blood samples were obtained at the time of admission to delivery from patients with and without preeclampsia. The diagnosis of preeclampsia was based on standard criteria as outlined in the Technical Bulletin from the American College of Obstetrics and Gynecology on hypertension in Pregnancy.7 The control group was composed of patients not meeting criteria for preeclampsia, as well as healthy controls. The study was approved by the institutional review board and all patients gave written informed consent. Table

Baseline Characteristics Troponin-I Levels

Age (years) Gestational age (weeks) C-section (n) Systolic BP (mm Hg) Diastolic BP (mm Hg) Troponin I (ng/mL)

Preeclampsia (n ⫽ 20)

Controls (n ⫽ 19)

28.2 36.5 9 160.4 88 0.008

24.6 39.2 7 137 81 0.01

P Value .055 .05 ⬍.01 .027 .59

819.e14

The American Journal of Medicine, Vol 120, No 9, September 2007

Cardiac Troponin I measurements were made on the Access II immunoassay analyzer using the AccuTnI method (Beckman Coulter, Inc., Fullerton, Calif). Statistical analysis was performed by sample T-Test using SPSS (Chicago, Ill) software.

RESULTS There were 39 women included (20 with preeclampsia and 19 without). Baseline characteristics of the 2 groups are summarized in the Table. Mean troponin I level was 0.008 ng/mL in patients with preeclampsia and 0.01 ng/mL in controls (P ⫽ .59) (Table). The highest troponin I level was 0.04 ng/mL for patients both with and without preeclampsia, occurring in only 1 patient in each group.

In conclusion, preeclampsia is not associated with an increase in troponin I levels. In patients with preeclampsia and elevated troponin levels, the troponin elevation should not be solely attributed to the inflammatory state of preeclampsia, and those patients should have further cardiac investigations and close cardiovascular follow-up.

CLINICAL SIGNIFICANCE ●

Preeclampsia is not associated with an increase in troponin I levels.

References

1. Bakshi TK, Choo MK, Edwards CC, Scott AG, Hart HH, Armstrong GP. Causes of elevated troponin I with a normal coronary angiogram. Intern ● In patients with preeclampsia and elevated Med J. 2002;32:520-525. troponin levels, the troponin elevation 2. Guest TM, Ramanathan AV, Tuteur should not be solely attributed to the inflamPG, Schechtman KB, Ladenson JH, Jaffe AS. Myocardial injury in critimatory state of preeclampsia, and those pacally ill patients. A frequently unrectients should have further cardiac investigaognized complication. JAMA. 1995; tions and close cardiovascular follow-up. 273:1945-1949. 3. Redman CW, Sacks GP, Sargent IL. Preeclampsia: an excessive maternal inflammatory response to pregnancy. Am J Obstet Gynecol. 1999;180:499-506. DISCUSSION 4. Roberts JM, Taylor RN, Goldfien A. Clinical and biochemical eviDespite hypertension and the presence of an inflammatory dence of endothelial cell dysfunction in the pregnancy syndrome state, troponin I was not found to be elevated in patients preeclampsia. Am J Hypertens. 1991;4:700-708. with preeclampsia. Although preterm delivery in a mother 5. Funai EF, Friedlander Y, Paltiel O, Tiram E, Xue X, Deutsch L, Harlap S. with preeclampsia is predictive of future late cardiovascular Long-term mortality after preeclampsia. Epidemiology. 2005;16:206-215. 6. Arnadottir GA, Geirsson RT, Arngrimsson R, Jonsdottir LS, Olafsson death,8 and population-based studies have shown increased O. Cardiovascular death in women who had hypertension in preglong-term maternal risk with hypertension related to pregnancy: a case-control study. Br J Obstet Gynaecol. 2005;112:286-292. nancy,6,9 the lack of troponin elevation at the time of de7. American College of Obstetrics and Gynecology. Hypertension in livery in our patient series does not support the occurrence pregnancy. The College; ACOG Tech Bull. No. 219; 1996. of acute myocardial damage. This is contrary to the only 8. Irgens HU, Reisaeter L, Irgens LM, Lie RT. Long term mortality of published report, where 6 patients with preeclampsia were mothers and fathers after pre-eclampsia: population based cohort study. BMJ. 2001;323:1213-1217. found to have mild elevation in cardiac troponin (median 9. Wilson BJ, Watson MS, Prescott GJ, et al. Hypertensive diseases of value 0.155 ng/mL).10 The cause of this discrepancy is unpregnancy and risk of hypertension and stroke in later life: results from clear and may be due to different patient characteristics such as cohort study. BMJ. 2003;326:845. concomitant systemic illnesses or intravascular volume status. 10. Fleming SM, O’Gorman T, Finn J, Grimes H, Daly K, Morrison JJ. Preeclampsia produces diffuse endothelial dysfunction as Cardiac troponin I in pre-eclampsia and gestational hypertension. Br J evidenced by increased levels of fibronectin, factor VIII antiObstet Gynaecol. 2000;107:1417-1420. 11. Friedman SA, Schiff E, Emeis JJ, Dekker GA, Sibai BM. Biochemical gen, thrombomodulin,11,12 endothelin, and thromboxanes and corroboration of endothelial involvement in severe preeclampsia. Am J decreased nitric oxide production. There is impaired flowObstet Gynecol. 1995;172:202-203. mediated vasodilation13,14 and impaired acetylcholine medi12. Hsu CD, Iriye B, Johnson TR, Witter FR, Hong SF, Chan DW. 15 ated vasorelaxation in preeclampsia when compared with Elevated circulating thrombomodulin in severe preeclampsia. Am J normal pregnancy. Whether the endothelial dysfunction is part Obstet Gynecol. 1993;169:148-149. of the cause of preeclampsia or simply a manifestation of the 13. McCarthy AL, Woolfson RG, Raju SK, Poston L. Abnormal endothelial cell function of resistance arteries from women with preeclampsia. disease is currently unknown. Regardless, the dysfunctional Am J Obstet Gynecol. 1993;168:1323-1330. endothelium persists for years after delivery16 and may explain 14. Cockell AP, Poston L. Flow-mediated vasodilatation is enhanced in the poorer cardiovascular outcomes. normal pregnancy but reduced in preeclampsia. Hypertension. 1997; Elevated troponin levels are often seen in a variety of 30:247-251. noncardiac conditions. Regardless of causes and despite ab15. Pascoal IF, Lindheimer MD, Nalbantian-Brandt C, Umans JG. Presence of acute coronary syndromes, troponin elevation renders eclampsia selectively impairs endothelium-dependent relaxation and leads to oscillatory activity in small omental arteries. J Clin Invest. a worse prognosis.17 Decision to pursue cardiac investigation 1998;101:464-470. relies on pretest probability of significant cardiac disease and 16. Chambers JC, Fusi L, Malik IS, Haskard DO, De Swiet M, Kooner JS. understanding of the pathophysiology of the primary noncarAssociation of maternal endothelial dysfunction with preeclampsia. diac illness. As preeclampsia is associated with increased risk JAMA. 2001;285:1607-1612. 5,6 of cardiovascular events, positive troponin levels should be 17. Ammann P, Maggiorini M, Bertel O, et al. Troponin as a risk factor for seen as a warning sign and not presumed elevated from the mortality in critically ill patients without acute coronary syndromes. J Am Coll Cardiol. 2003;41:2004-2009. inflammatory state.

The American Journal of Medicine (2007) 120, e1-e2

CLINICAL COMMUNICATION TO THE EDITOR Rapidly Reversible Cardiogenic Shock as a Pheochromocytoma Presentation To the Editor: A 52-year-old woman presented to the emergency department with a 12-hour history of fever, nausea, emesis, and diffuse abdominal pain. She was found to have mild abdominal distention with diminished bowel sounds, leukocytosis, and hyperglycemia. Within several hours, she developed progressively severe cardiogenic shock and required endotracheal intubation for worsening respiratory distress due to acute pulmonary edema. Her electrocardiogram revealed ST depression in the anterolateral leads, and laboratory values included a CPK of 710 U/L, a CK-MB of 47.4 ng/mL (6.7%), and a troponin I of ⬎50 ng/mL. Limited transthoracic echocardiography demonstrated mild dilation of the left ventricle with severe systolic dysfunction and an ejection fraction of 25% to 30%. The entire base of the heart and the mid-portion of the left ventricle were nearly akinetic, whereas the left ventricular apex contracted vigorously. An emergent left heart catheterization revealed normal coronary arteries and severe global systolic left ventricular dysfunction with an ejection fraction of 30%. The patient was hemodynamically stabilized on an intraaortic balloon pump and a milrinone infusion. A transesophageal echocardiogram on the following day showed only mild hypokinesis at the posterior and septal walls with an ejection fraction of 65%. Abdominal computed tomography (CT) and magnetic resonance imaging (MRI) examinations revealed a left adrenal tumor, suggestive of a pheochromocytoma (Figures 1 and 2). The diagnosis of an adrenal pheochromocytoma was confirmed by the detection of significantly elevated plasma and urine catecholamines, with epinephrine levels 10 to 20 times above the upper limit of normal in plasma and urine. The patient returned several weeks after recovery for a laparoscopic adrenalectomy, following alpha-adrenergic blockade with phenoxybenzamine, as well as subsequent beta-adrenergic blockade for resting tachycardia, hypertension, and nonsustained ventricular tachycardia, and metyrosine therapy to Requests for reprints should be addressed to Kirsten Brukamp, MD, Renal-Electrolyte and Hypertension Division, Department of Medicine, University of Pennsylvania School of Medicine, 700 Clinical Research Building, 415 Curie Boulevard, Philadelphia, PA 19104. E-mail address: [email protected]

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.

Figure 1 The patient’s abdominal computed tomography (CT) examination demonstrates a left adrenal mass. The axial contrastenhanced CT image through the upper abdomen with oral and intravenous contrast reveals a well-circumscribed mildly heterogeneous 3.2 cm mass (arrow) within the left adrenal gland, which did not contain fat. Major differential diagnostic considerations include adrenal metastasis, pheochromocytoma, and adenoma. There was no demonstrable gastrointestinal pathology, and the intraabdominal structures, such as aorta, liver, spleen, pancreas, and kidneys, appeared normal on cross-sectional imaging.

inhibit catecholamine synthesis. The surgically extirpated mass was identified as a pheochromocytoma by histopathologic examination. Not all patients with pheochromocytoma exhibit the classic findings,1 namely persistent or paroxysmal hypertension, headache, visual blurring, tachycardia, diaphoresis, and pallor. Many have cardiovascular and endocrine symptoms, such as orthostatic hypotension, cardiomyopathy, flushing, weight loss, hyperglycemia, and neuropsychiatric changes, which may still occur with a frequency of ⬎10% in patients.1 Although accompanying nonspecific nausea and vomiting are fairly common,1 distinct abdominal and gastrointestinal manifestations, albeit rarely described, can lead to a dramatic presentation as part of a grave acute abdominal pain syndrome. This patient exhibited acute myocardial dysfunction as the main acute complication, which led to life-threatening cardiogenic shock and acute pulmonary edema, with rapid reversibility. The various cardiac manifestations of pheochromocytomas can be classified as cardiomyopa-

e2

The American Journal of Medicine, Vol 120, No 9, September 2007 pheochromocytomas. Acutely life-threatening cardiovascular or severe abdominal complications of pheochromocytomas may pose a diagnostic and therapeutic challenge in the intensive care setting or in the emergency department, with potential medicolegal implications.4 Kirsten Brukamp, MD Simin Goral, MD Raymond R. Townsend, MD Renal-Electrolyte and Hypertension Division Department of Medicine University of Pennsylvania School of Medicine Philadelphia

Frank E. Silvestry, MD Cardiovascular Medicine Division Department of Medicine University of Pennsylvania School of Medicine Philadelphia

Figure 2 A heterogeneous, non-lipid-containing left adrenal mass is confirmed on magnetic resonance imaging (MRI). The axial heavily T2-weighted MR image through the upper abdomen, performed after the CT scan, demonstrates areas of mixed intermediate and high signal intensity within the non-lipid-containing adrenal mass (arrow). Chemical shift imaging did not reveal the presence of microscopic lipid (not shown). A heterogeneous T2weighted signal intensity and arterial phase enhancement were suggestive of a pheochromocytoma.

Drew A. Torigian, MD, MA Department of Radiology University of Pennsylvania School of Medicine Philadelphia

doi:10.1016/j.amjmed.2006.08.031 thies, forms of myocardial ischemia, and primarily arrhythmogenic conditions. Takotsubo cardiomyopathy2 also is known as transient left ventricular apical ballooning or stress-induced cardiomyopathy, and it is defined as a reversible dysfunction of the left ventricular apex with compensatory hyperkinesis of the basal walls, resulting in systolic apical ballooning in the setting of absent coronary artery disease. Takotsubo cardiomyopathy has been described in pheochromocytoma, and the patient presented here is only the second one in whom a rapidly reversible inverted takotsubo pattern has been noted.3 The patient’s hospital course demonstrates the variability of clinical presentations that may be associated with adrenal

References 1. Lenders JW, Eisenhofer G, Mannelli M, Pacak K. Phaeochromocytoma. Lancet. 2005;366:665-675. 2. Bybee KA, Kara T, Prasad A, et al. Systematic review: transient left ventricular apical ballooning: a syndrome that mimics ST-segment elevation myocardial infarction. Ann Intern Med. 2004; 141:858-865. 3. Sanchez-Recalde A, Costero O, Oliver JM, et al. Images in cardiovascular medicine. Pheochromocytoma-related cardiomyopathy: inverted Takotsubo contractile pattern. Circulation. 2006;113:e738-e739. 4. Preuss J, Woenckhaus C, Schwesinger G, Madea B. Non-diagnosed pheochromocytoma as a cause of sudden death in a 49-year-old man: a case report with medico-legal implications. Forensic Sci Int. 2006;156: 223-228.

The American Journal of Medicine (2007) 120, e3

CLINICAL COMMUNICATION TO THE EDITOR Cold Water-induced Pulmonary Edema To the Editor: A 54-year-old white woman with a history of hypertension experienced shortness of breath without chest pain or palpitations after swimming in relatively cold water (22°C) while the ambient temperature was 43°C. Paramedics found her to be in pulmonary edema with a blood pressure of 150/100 mm Hg and low oxygen saturation (SO2 ⫽ 64%). She was treated with supplemental oxygen and furosemide with slight improvement and later was treated in the emergency department with continuous positive airway pressure. An electrocardiogram revealed minimal ST depression in leads I and aVL. Troponin I was elevated (11 ␮g/L). On echocardiography, global severe left ventricular (LV) and right ventricular (RV) dysfunction was noticed. Her condition improved following supportive therapy. Angiography, performed on day 4, demonstrated normal coronary arteries and LV function, with no evidence of renal artery stenosis. Echocardiography, on day 5, revealed normal LV and RV size and function, with no evidence of valvular pathology. Both cardiac magnetic resonance imaging and endothelial function were normal. She was discharged on day 7.

The underlying mechanisms of cold water-induced pulmonary edema remain elusive. The most common explanation is that the decrease in core body temperature results in the redistribution of blood from peripheral to thoracic vessels. This occurs through vasoconstriction that triggers an increase in both preload and afterload. Decrease in core body temperature also has been shown to increase pulmonary vascular resistance. Immersion also increases sympathetic activity, which among other effects causes peripheral vasoconstriction.4 All these factors seem sufficient to rupture pulmonary capillary membranes, resulting in pulmonary edema. Normal echocardiograms in most patients could be explained by delayed testing. Our case demonstrates LV dysfunction might be transient. Pulmonary edema as a result of swimming in cold water is a rare and understated condition. Clinicians should be aware of it to provide optimal medical management. Roy Beinart, MD Shlomi Matetzky, MD The Heart Institute, Sheba Medical Center, Tel Hashomer Sackler Faculty of Medicine Tel Aviv University, Israel

Tova Arad, MD Department of Internal Medicine Yoseftal Medical Center, Eilat, Israel

DISCUSSION Acute pulmonary edema has been described in swimmers and divers, but its prevalence is unknown. Because mild cases might be unrecognized clinically, it has more than likely been under-reported.1-6 The majority of cases resolve with supportive treatment.4 Koehle et al3 described 60 cases, of whom most were men (66%). Swimming accounted for one third of the cases, primarily in cold water temperatures (mean of 15.2°C), but some occurred in water as warm as 29°C. The most frequently performed investigation was chest radiography, demonstrating pulmonary infiltrates that resolved after treatment. Electrocardiography exhibited minor, nonspecific changes,3 while few patients had positive troponin levels.1 Several patients were assessed by echocardiography, but only 2 had documented LV dysfunction.2,6 Requests for reprints should be addressed to Roy Beinart, MD, Heart Institute, Sheba Medical Center, Tel Hashomer 52621, Israe. E-mail address: [email protected]

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.

Hanoch Hod, MD The Heart Institute, Sheba Medical Center, Tel Hashomer Sackler Faculty of Medicine Tel Aviv University, Israel

doi:10.1016/j.amjmed.2006.08.030

References 1. Biswas R, Shibu PK, James CM. Pulmonary oedema precipitated by cold water swimming. Br J Sports Med. 2004;38:e36. 2. Cochard G, Arvieux J, Lacour JM, Madouas G, Mongredien H, Arvieux CC. Pulmonary edema in scuba divers: recurrence and fatal outcome. Undersea Hyperb Med. 2005;32:39-44. 3. Koehle MS, Lepawsky M, McKenzie DC. Pulmonary oedema of immersion. Sports Med. 2005;35:183-190. 4. Lund KL, Mahon RT, Tanen DA, Bakhda S. Swimming-induced pulmonary edema. Ann Emerg Med. 2003;41:251-256. 5. Slade JB Jr, Hattori T, Ray CS, Bove AA, Cianci P. Pulmonary edema associated with scuba diving: case reports and review. Chest. 2001;120:16861694. 6. Wilmshurst PT, Nuri M, Crowther A, Webb-Peploe MM. Cold-induced pulmonary oedema in scuba divers and swimmers and subsequent development of hypertension. Lancet. 1989;1:62-65.

The American Journal of Medicine (2007) 120, e5-e6

CLINICAL COMMUNICATION TO THE EDITOR Unexplained Isolated Elevation in Serum Aspartate Aminotransferase: Think Macro! To the Editor: In patients with elevated aminotransferase levels without liver disease, the following conditions should be considered. False-positive elevations in aspartate aminotransferase (AST) may be seen in patients receiving erythromycin estolate,1 with primary muscle disease, with celiac disease, with thyroid disease, with adrenal insufficiency, and with hemolytic disorders. Rare cases have been described of moderate elevations in plasma AST levels caused by the presence of “macro”-AST in which AST complexes with immunoglobulin.2 A healthy 60-year-old woman was referred to our liver center for evaluation of a persistently elevated blood AST level (Table 1). Her medical history included mitral valve prolapse and hyperlipidemia, for which she took no medications. There was no personal or family history of jaundice, skeletal muscle, or thyroid or hemolytic disorders. Physical examination was unrevealing. Prior workup had included a normal abdominal computed tomography scan and negative serologies for viral, autoimmune, and metabolic liver diseases. All of the nonhepatic causes of AST elevation, as described above, were ruled out. Electrophoresis testing confirmed that she had an immunoglobulin-comRequests for reprints should be addressed to David A. Sass, MD, Assistant Professor of Medicine, Director of Hepatology Fellowship Program and Associate Director of GI Fellowship, Division of Gastroenterology, Hepatology and Nutrition, University of Pittsburgh School of Medicine, Kaufmann Medical Building, Suite 916, 3471 Fifth Avenue, Pittsburgh, PA 15213. E-mail address: [email protected]

Table 1

plexed AST. This macromolecule presumably caused a persistent increase in measured AST activity. Several enzymes complexed with immunoglobulins have been described in the literature.3 This group of disorders was termed “immunoglobulin-complexed enzyme” disorders by Klonoff.4 Macroamylasemia is the most frequently reported, but creatine kinase, lactate dehydrogenase, alkaline phosphatase, alanine aminotransferase, glucose-6-phosphate dehydrogenase, and AST are other examples. AST, which may be cytoplasmic or mitochondrial, is a well-known enzyme found in clinically significant amounts in the tissue of liver, skeletal muscle, heart, and, to a lesser extent, erythrocytes. Although the majority of macro-AST cases have been described in asymptomatic individuals,5,6 a number of cases have been described in association with acute and chronic hepatitis,7,8 autoimmune hepatitis, inflammatory bowel disease,9 and hepatic and other gastrointestinal malignancies, and after specific allergen injection immunotherapy.10 Although the pathogenesis of macroenzyme formation remains speculative, an autoimmune process has been proposed as the likely mechanism.3 The presence of macro-AST can be determined by way of exclusion chromatography, electrophoresis, and activation assays with pyridoxal 5-phosphate. AST levels in patients with macroAST have been shown to fluctuate over time and may, in fact, be normal at times.11 An appreciation of macro-AST may obviate the need for more invasive investigations to which patients may be unnecessarily subjected. It also is important to recognize it as a benign condition for which patients can be reassured that no specific treatment is required. David A. Sass, MD Rajagopal Chadalavada, MD

Trend of Liver Enzymes

Date

Total Bilirubin (0.3-1.5)* mg/dL

AST (15-37)* IU/L

ALT (15-37)* IU/L

Alkaline Phosphatase (40-125)* IU/L

GGT (⬍65)* IU/L

Albumin (3.4-5.0)* g/dL

08/13/05 12/06/05 03/22/06 03/30/06 04/26/06

0.4 0.18 0.19 0.26 0.3

427 363 368 399 427

28 30 37 37 24

109 94 123 134 123

27 8 10 12 16

4.2 3.6 3.8 4.1 4.7

AST ⫽ aspartate aminotransferase; ALT ⫽ alanine aminotransferase; GGT ⫽ gamma-glutamyl transferase. *Reference ranges.

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.

e6

The American Journal of Medicine, Vol 120, No 9, September 2007 Division of Gastroenterology, Hepatology and Nutrition Department of Medicine

Mohamed A. Virji, MD, PhD Division of Clinical Chemistry, Department of Pathology University of Pittsburgh, Penn

doi:10.1016/j.amjmed.2006.10.022

References 1. Sabath LD, Gerstein DA, Finland M. Serum glutamic oxaloacetic transaminase. False elevations during administration of erythromycin. N Engl J Med. 1968;279:1137. 2. Litin SC, O’Brien JF, Pruett S, et al. Macroenzyme as a cause of unexplained elevation of aspartate aminotransferase. Mayo Clin Proc. 1987;62:681. 3. Remaley AT, Wilding P. Macroenzymes: biochemical characterization, clinical significance, and laboratory detection. Clin Chem. 1989; 2261-2270. 4. Klonoff DC. Macroamylasemia and other immunoglobulin-complexed enzyme disorders. West J Med. 1980;133:392-407.

5. Weidner N, Lott JA, Yale VD, Wahl RL, Little RA. Immunoglobulin-complexed aspartate aminotransferase. Clin Chem. 1983;29: 382-384. 6. Wiltshire EJ, Crooke M, Grimwood K. Macro-AST a benign cause of persistently elevated aspartate aminotransferase. J Pediatr Child Health. 2004;40:642-643. 7. Collins J, Ritter D, Bacon BR, et al. Macro-aspartate aminotransferase in a female with antibodies to hepatitis C virus. Liver. 2002;22:501-506. 8. Kanemitsu F. Aspartate aminotransferase-immunoglobulin G complexes in sera of two patients with acute hepatitis and immunoblastic lymphadenopathy. Clin Chim Acta. 1986;161:353-354. 9. Tajiri H, Nakano T, Kozaiwa K, et al. Immunoglobulin-complexed aspartate aminotransferase with a possible association with ulcerative colitis and its activity. J Clin Lab Immunol. 1992;38:41-49. 10. Triester S, Douglas DD. Development of macro-aspartate aminotransferase in a patient undergoing specific allergen injection immunotherapy. Am J Gastroenterol. 2005;100:243-245. 11. Orlando R, Carbone A, Lirussi F. Macro-aspartate aminotransferase (macro-AST). A 12-year follow-up study in a young female. Eur J Gastroenterol Hepatol. 2003;15:1371-1373.

The American Journal of Medicine (2007) 120, e7-e9

CLINICAL COMMUNICATIONS TO THE EDITOR Cinacalcet-Associated Graft Dysfunction and Nephrocalcinosis in a Kidney Transplant Recipient To the Editor: Cinacalcet (Sensipar; Amgen Inc., Thousand Oaks, Calif) is a calcimimetic agent that potentiates the action of extracellular Calcium (Ca) on the calcium-sensing receptor and has been shown to be effective in treating both primary and secondary hyperparathyroidism.1,2 In a recent clinical trial, cinacalcet has been shown to be superior in reducing intact parathyroid hormone (iPTH), Ca, phosphorous (P), and the Ca*P in end-stage renal disease patients compared with conventional therapy using vitamin D analogues alone.1 The major adverse effects reported include nausea, vomiting, diarrhea, and hypocalcemia. Management with cinacalcet of hyperparathyroidism that persists following renal transplantation has not been systematically studied. We report here the first case suggesting that cinacalcet may contribute to the development of acute renal failure and nephrocalcinosis in the setting of posttransplant hypercalcemia and hyperparathyroidism.

CASE REPORT A 53-year-old woman with end-stage renal disease from unknown etiology underwent cadaveric renal transplantation. A biopsy of her native kidney showed advanced chronic tubulointerstitial changes without Ca deposition. She was managed with chronic hemodialysis for 5 years before the transplant. During that time she developed secondary hyperparathyroidism that had been largely treated with doxercalciferol. Treatment with cinacalcet, 30 mg daily, was added 1 month before transplantation, because of inadequate control of hyperparathyroidism in association with hypercalcemia and hyperphosphatemia (bio-intact PTH [biPTH] 497 pg/mL, Ca 10.6 mg/dL, and P 7.4 mg/ dL). At the time of transplantation, her Ca and P were 10.9 mg/dL and 7.2 mg/dL, respectively. Kidney transplantation was performed without any immediate complications, and she had an uneventful postopRequests for reprints should be addressed to Yeong-Hau Howard Lien, MD, PhD, Department of Medicine, University of Arizona, 1501 N. Campbell Ave., Tucson, AZ 85724. E-mail address: [email protected]

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.

erative course. Her immunosuppression therapy consisted of FK506, mycophenolate mofetil, and prednisone. Figure 1 summarizes the changes in renal function and Ca, P, FK506, and iPTH levels in relationship with administration of cinacalcet. Cinacalcet therapy (30 mg per day) was added because of worsening hypercalcemia and hyperparathyroidism in the first few weeks after transplantation. The treatment course spanned day 17 through 61, with the exception of days 27 to 32. This interruption was the result of a brief hospitalization for diarrhea, which resolved after hydration and reduction of the mycophenolate mofetil dose. Cinacalcet was held during the hospitalization only because it was unavailable in the inpatient pharmacy. Her serum creatinine (Cr) increased to 1.9 mg/dL after resuming cinacalcet. She was euvolemic with stable weight and blood pressure at the time. The FK506 trough level was 8.8 ng/mL, and her graft ultrasound examination showed no hydronephrosis. She underwent a graft biopsy on day 44 that revealed Ca deposition in the tubules with surrounding mononuclear cell infiltration and fibrosis (Figure 2). There was no pathological evidence of rejection, viral infection, or a variety of conditions associated with nephrocalcinosis, including acute tubular necrosis, sarcoidosis, and calcineurin inhibitor toxicity. In addition, the donor was 20 years old at the time of death and the mate kidney was functioning well. It is unlikely that pathological changes were due to de novo nephrocalcinosis. Testing of her serum for cytomegalovirus (CMV) by DNA hybridization assay and for Polyoma BK virus by reverse transcription polymerase chain reaction also was negative. Measurement of her 24-hour urine Ca yielded 317 mg/day (normal in females up to 250 mg/day) and 370 mg/day measured on day 52 and day 59, respectively. Further evidence of the hypercalciuria was noted on day 55 when gross calcification around the ureteral stent of the graft was identified without obstruction. The patient’s serum Cr level peaked at 2.7 mg/dL on day 56, and cinacalcet was discontinued on day 61. Her serum Ca and iPTH increased rapidly after stopping cinacalcet but her serum Cr gradually decreased (Figure 1). Repeat determination of urine Ca excretion on day 69 was normal at 176 mg/day. A subtotal parathyroidectomy was performed on day 84 for persistent hypercalcemia and hyperparathyroidism. Subsequently, her Ca, P, and iPTH levels decreased as predicted. She had a transient increase of serum Cr postparathyroidectomy that eventually stabilized at 1.6 mg/dL.

e8

The American Journal of Medicine, Vol 120, No 9, September 2007

Figure 1 The post-transplant time course of the changes in serum Cr, Ca, P, FK506, and iPTH levels in relationship with administration of cinacalcet. The bottom black bar depicts the presence of cinacalcet and the arrowhead depicts the subtotal parathyroidectomy. All units are in mg/dL except for iPTH, which is in pg/mL, and FK506 in ng/mL. The iPTH was determined by a non-third-generation electro-chemiluminescent immunoassay (reference range 15-75 pg/mL) except for that obtained on day 204, at which time a biPTH chemiluminescent assay was used (reference range 6-40 pg/mL). To convert biPTH to iPTH, we multiply by 2.27.4

At the last follow-up (day 204), her Cr, Ca, and P were 1.6, 8.5, and 3.7 mg/dL, respectively, and biPTH was 22 pg/mL.

DISCUSSION In the presence of other contributing factors, such as resorption of soft tissue calcification, increased calcitriol synthesis from renal graft, severe hyperparathyroidism, and immunosuppression therapy including FK506 and prednisone, we believe that the administration of cinacalcet may have triggered the events leading to graft dysfunction and nephrocalcinosis in our case. The withdrawal of cinacalcet alone lowered serum Cr and urine Ca levels. This temporal correlation among graft dysfunction, hypercalciuria, and the usage of cinacalcet supports strongly the suggestion that cinacalcet plays a key role. It is less likely that the nephrocalcinosis was a result of hyperparathyroidism per se because the urinary Ca excretion was the greatest when the iPTH was lowered by cinacalcet. Additionally, the increase

in serum Cr occurred as iPTH and serum Ca were suppressed. It also is unlikely that the graft dysfunction was due to hypercalcemia because the serum Ca levels were below the range of 12-15 mg/dL, which has been reported to cause post-transplant renal failure.3,4 After kidney transplantation, there is resorption of soft tissue Ca phosphate deposits and increased calcitriol production that further contributes to the positive Ca balance by increasing gastrointestinal absorption.5 Severe hyperparathyroidism also played a supportive role in our case. In the presence of cinacalcet, the patient’s iPTH levels were significantly elevated, which increases bone resorption and, consequently, renal Ca load. Calcineurin inhibitors and glucocorticoids also enhance bone resorption to increase renal Ca load.4 In addition, calcineurin inhibitors may increase renal calcium excretion by downregulating calcium binding protein, Calbindin D28K, in the distal tubule.6 The addition of cinacalcet in this setting is likely to markedly increase urine calcium excretion, be-

Peng et al

Cinacalcet and Graft Dysfunction and Nephrocalcinosis

e9 Luon W. Peng, DO Joy L. Logan, MD Sam H. James, MD Orthopedic Research Lab Department of Medicine University of Arizona Tucson

Katherine M. Scott, MD Department of Pathology University of Arizona Tucson

Yeong-Hau Howard Lien, MD, PhD Orthopedic Research Lab Department of Medicine University of Arizona Tucson

Figure 2 Histopathology of renal graft biopsy with the arrows showing tubular destruction with intratubular calcium deposition (H&E, original magnification ⫻400).

cause activation of calcium-sensing receptor in the cortical thick ascending limb7 would block paracellular calcium transport, a similar scenario as in the autosomal dominant familial hypocalcemia due to gain-of-function mutations of the calcium-sensing receptor.8 Recently, Serra et al reported that cinacalcet was effective in lowering hypercalcemia in 11 renal transplant recipients with hyperparathyroidism.9 They did not observe graft dysfunction or hypercalciuria.9 We think that the major differences between their patients and ours are that theirs had mild hyperparathyroidism with a mean iPTH and serum Ca of 176 pg/mL and 10.9 mg/dL, respectively, and that the treatment of cinacalcet was much later after kidney transplantation (6-364 months). Hypercalciuria and renal dysfunction have not been reported in cinacalcet-treated patients with primary hyperparathyroidism—whose iPTH levels are much lower than our patient— either.2 Therefore, we would like to point out that cinacalcet can be used in renal transplant recipients; however, we caution that early use of cinacalcet in the presence of severe hyperparathyroidism and intense immunosuppression may be risky. Our case underscores the necessity for further study to determine the safety of using cinacalcet for hyperparathyroidism in renal transplant recipients.

doi:10.1016/j.amjmed.2005.09.041

References 1. Block GA, Martin KJ, de Francisco AL, et al. Cinacalcet for secondary hyperparathyroidism in patients receiving hemodialysis. N Engl J Med. 2004;350:1516-1525. 2. Peacock M, Bilezikian JP, Klassen PS, Guo MD, Turner SA, Shoback D. Cinacalcet hydrochloride maintains long-term normocalcemia in patients with primary hyperparathyroidism. J Clin Endocrinol Metab. 2005;90:135-141. 3. Lins LE. Renal function in hypercalcemia after renal transplantation. Scand J Urol Nephrol Suppl. 1977;(42):159-161. 4. Torres A, Lorenzo V, Salido E. Calcium metabolism and skeletal problems after transplantation. J Am Soc Nephrol. 2002;13:551-558. 5. Julian BA, Quarles LD, Niemann KM. Musculoskeletal complications after renal transplantation: pathogenesis and treatment. Am J Kidney Dis. 1992;19:99-120. 6. Lee CT, Huynh VM, Lai LW, Lien YH. Cyclosporine A-induced hypercalciuria in calbindin-D28k knockout and wild-type mice. Kidney Int. 2002;62:2055-2061. 7. Riccardi D, Hall AE, Chattopadhyay N, Xu JZ, Brown EM, Hebert SC. Localization of the extracellular Ca2⫹/polyvalent cation-sensing protein in rat kidney. Am J Physiol. 1998;274:F611-F622. 8. Pidasheva S, D’Souza-Li L, Canaff L, Cole DE, Hendy GN. CASRdb: calcium-sensing receptor locus-specific database for mutations causing familial (benign) hypocalciuric hypercalcemia, neonatal severe hyperparathyroidism, and autosomal dominant hypocalcemia. Hum Mutat. 2004;24:107-111. 9. Serra AL, Schwarz AA, Wick FH, Marti HP, Wuthrich RP. Successful treatment of hypercalcemia with cinacalcet in renal transplant recipients with persistent hyperparathyroidism. Nephrol Dial Transplant. 2005; 20:1315-1319.

The American Journal of Medicine (2007) 120, e11-e12

CLINICAL COMMUNICATION TO THE EDITOR Q Fever in US Military Returning from Iraq To the Editor: Q fever is a zoonotic infection caused by Coxiella burnetii that may present with acute or chronic clinical manifestations, including a nonspecific febrile illness, pneumonia, hepatitis, or endocarditis.1 Although unusual in the US, Q fever is endemic in the Middle East. Our case series raises the concern that Q fever may be a significant infectious disease threat to US military serving in Iraq.2 In November 2004, 2 military service members from the same Marine battalion became ill upon return from a 7-month tour of duty in Iraq. The first patient was a 35-year-old medical corpsman who experienced fever to 39.4°C, chills, rigors, severe retro-orbital headache, loose stools, and lower back and calf pain. Cerebrospinal fluid analysis was normal. A chest radiograph and an abdominal computed tomography scan with intravenous and oral contrast also were normal. Although the patient defervesced without antibiotics, his serology for Q fever subsequently returned positive, and he was treated with 21 days of doxycycline. The second patient, a 23-year-old US Marine infantryman, experienced fever to 41.1°C, chills, rigors, sore throat, headache, and diffuse myalgias. His chest radiograph demonstrated diffuse bilateral interstitial markings. He defervesced within 24 hours after starting empiric doxycycline, and his Q fever serology later returned positive. Both soldiers noted goats wandering through the camp but denied other exposures. More recently, in September 2006, a third patient, a 24year-old US Marine officer, developed fevers to 39.3°C, chills, headache, and photophobia 3 weeks after returning from Iraq. He denied specific exposures. His chest radiograph was normal, as was analysis of his cerebrospinal spinal fluid. He defervesced within 24 hours of initiating empiric doxycycline. His serology was negative for Q fever on admission but was positive when repeated 1 week later. All 3 patients had similar laboratory abnormalities including

The opinions and assertions contained herein are those of the authors and are not to be construed as official or as reflecting the views of the Department of Defense, the Department of the Navy, or the naval services at large. Requests for reprints should be addressed to Todd D. Gleeson, MD, Division of Infectious Diseases, Department of Internal Medicine, National Naval Medical Center, 8901 Wisconsin Avenue, Bethesda, MD 20889. E-mail address: [email protected]

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.

hyponatremia, hepatitis, leucopenia, and thrombocytopenia (Table). In each case, the serology for acute Q fever was positive by an immunofluorescence assay (Focus Technologies Q Fever IFA IgM and IgG assays). Interestingly, the first 2 patients had an isolated positive hepatitis B core antibody with other hepatitis B studies negative. Repeat hepatitis B core antibody testing was negative on discharge. In addition to negative cultures of blood, urine, stool, and cerebrospinal fluid, other negative studies included stool and throat swabs for Enterovirus by PCR, malaria smears, serologies for dengue, sand fly fever viruses, Rift Valley Fever, West Nile virus, Leptospira, Legionella, Bartonella, Borrelia, Brucella, Ehrlichia, and several Rickettsia species. Although it is endemic in Iraq, only 3 cases of Q fever were reported during the Gulf War (1990-1991).3 In 2003, an evaluation of 62 cases of pneumonia in Iraq-deployed US troops revealed 8 Q fever seroconversions,4 and Leung-

Table Selected Laboratory Values and Serological Test Results for Q Fever* Laboratory (normal range)

Patient 1

Patient 2

Patient 3

Serum sodium (137-145 mEq/L) Aspartate transaminase (17-49 Units/L) Alanine transaminase (7-56 Units/L) Alkaline phosphatase (36-126 Units/L) Total bilirubin (0.2-1.3 mg/dL) WBC (4.0-11.0 thousand cells/␮L) Platelets (150-450 thousand cells/ ␮L) Coxiella burnetii Phase I IgM (ⱕ1:64) Coxiella burnetii Phase I IgG (ⱕ1:16) Coxiella burnetii Phase II IgM (ⱕ1:64) Coxiella burnetii Phase II IgG (ⱕ1:16)

130

130

134

128

351

305

155

258

425

455

356

186

1.8

4.9

0.7

3.6

3.8

2.0

120

45

89

ⱖ1:1024

ⱖ1:1024

ⱖ1:1024

1:512

1:32

1:16

ⱖ1:1024

ⱖ1:1024

1:512

ⱖ1:1024

ⱖ1:1024

1:128

WBC ⫽ white blood cell. *Serology by Focus Diagnostics Q Fever Immunofluorescence Assay.

e12

The American Journal of Medicine, Vol 120, No 9, September 2007

Shea et al. recently reported 2 cases of acute Q fever in military service members.2 We postulate that our soldiers acquired C. burnetii infection via aerosolization of infectious particles most likely associated with the goats in the environment. In addition, although rheumatoid factor, antiphospholipid antibodies, and smooth muscle antibodies are common in patients with Q fever,1 a false-positive hepatitis B core antibody has not previously been described. Our case series raises the concern that Q fever may be occurring in our military at a higher rate than previously predicted. As Q fever has a long incubation period (average 14-21 days) and may present as acute or chronic disease, nonmilitary physicians should be aware of this potential diagnosis in individuals with this geographic exposure.

Mark D. Johnson, MD Department of Internal Medicine National Naval Medical Center Bethesda, Md

Joshua D. Hartzell, MD Division of Infectious Diseases Department of Internal Medicine Walter Reed Army Medical Center Washington, DC

John R. Mascola, MD Vaccine Research Center National Institute of Allergy and Infectious Diseases National Institutes of Health Bethesda, Md

doi:10.1016/j.amjmed.2007.03.020

References Todd D. Gleeson, MD Catherine F. Decker, MD Division of Infectious Diseases National Naval Medical Center Bethesda, Md

1. Maurin M, Raoult D. Q fever. Clin Microbiol Rev. 1999;12:518-553. 2. Leung-Shea C, Danaher PJ. Q fever in members of the United States armed forces returning from Iraq. Clin Infect Dis. 2006;43:e77-e82. 3. Ferrante MA, Dolan MJ. Q fever meningoencephalitis in a soldier returning from the Persian Gulf War. Clin Infect Dis. 1993;16:489-496. 4. Anderson AD, Smoak B, Shuping E, et al. Q fever and the US military. Emerg Infect Dis. 2005;11:1320-1322.

The American Journal of Medicine (2007) 120, e13

LETTER Treatment of Depression in Patients with Coronary Heart Disease To the Editor: Vieweg and colleagues provide excellent updates on treating depression in patients with coronary artery disease. Their practical approach, however, leaves out an important intermediate step. For the primary care physician with a large number of patients with possible depression, some confirmatory testing of patients identified on a 2-question screen would be appropriate. In my family practice residency program, we used the 2-question screen and subsequent test with Patient Health Questionnaire-9 as a screening tool for confirmation. The Patient Health Questionnaire-9 is a well-validated, convenient, and easy-toapply tool to screen depression and score the severity. It also has the best positive likelihood ratio and positive predictive value compared with the other commonly available screening instruments.1 I recommend that all primary care providers and even cardiologists use this convenient tool that can improve the care of patients with depression.2

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.

US Preventive Services Task Force recommendations state that “screening improves the accurate identification of depressed patients in primary care settings and that treatment of depressed adults identified in primary care settings decreases clinical morbidity.”3 Those patients with coronary artery disease can be seen, screened, and treated best at primary care settings. Hajime Kojima, MD Faculty Development Fellow University of Pittsburgh Department of Family Medicine Pittsburgh, Penn

doi:10.1016/j.amjmed.2006.07.037

References 1. Vieweg WVR, Julius DA, Fernandez A, et al. Treatment of depression in patients with coronary heart disease. Am J Med. 2006;119:567-573. 2. Nease DE Jr, Maloin JM. Depression screening: a practical strategy. J Fam Pract. 2003;52(2):118-124. 3. Gary Ruoff. A method that dramatically improves patient adherence to depression treatment. J Fam Pract. 2005;54(10):846-852. 4. Screening for depression. Release date: May 2002. U.S. Preventive Services Task Force. Available at: http://www.ahrq.gov/clinic/uspstf/ uspsdepr.htm. Accessed July 19, 2006.

The American Journal of Medicine (2007) 120, e15

LETTER The Reply: My colleagues and I appreciate Kojima’s kind remarks. We agree that the PHQ-9 questionnaire is an excellent tool to confirm a diagnosis of depression in a primary care setting. We refer readers interested in learning more about the PHQ-9 to: http://www.pfizer.com/pfizer/ download/do/phq-9.pdf#search⫽%22patient%20health%20

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.

questionnaire%20(PHQ-9)%22, or simply place the key words Patient Health Questionnaire (PHQ-9) into the Google search engine to be directed to the same site. W. Victor R. Vieweg, MD Department of Psychiatry Medical College of Virginia Richmond

doi:10.1016/j.amjmed.2006.09.028

The American Journal of Medicine (2007) 120, e17

LETTERS Spinal Cord Stimulation: A New Form of Pain Modulatory Treatment in Cardiac Syndrome X

Gregory Angelo Sgueglia, MD Alfonso Sestito, MD, PhD Istituto di Cardiologia Università Cattolica del Sacro Cuore Roma, Italy

To the Editor: We read with interest the review by Hurst et al on cardiac syndrome X (CSX) in The American Journal of Medicine.1 As correctly pointed out by Hurst et al, pain modulation is an important form of therapy in patients with CSX. However, authors have not taken in account a new form of pain modulatory treatment represented by spinal cord stimulation (SCS). Initially proposed to treat refractory angina pectoris in patients with obstructive coronary artery disease not suitable for revascularization,2 SCS is now included in the American College of Cardiology/American Heart Association guidelines on chronic stable angina SCS as a Class II recommendation for refractory angina.3 Moreover, several studies have shown that SCS is a very effective and safe treatment for patients with refractory CSX and determines improvement of symptoms and quality of life in patients with angina and normal coronary arteries or patients with typical CSX.4-6 In a crossover study, we recently documented favorable effects of SCS on angina episodes, as well as on episodes of ST-segment depression on Holter monitoring and on both symptoms and ST-segment changes induced by dobutamine stress test, in a small group of CSX patients treated with SCS.6

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.

doi:10.1016/j.amjmed.2006.09.024

References 1. Hurst T, Olson TH, Olson LE, Appleton CP. Cardiac syndrome X and endothelial dysfunction: new concepts in prognosis and treatment. Am J Med. 2006;119:560-566. 2. Mannheimer C, Eliasson T, Augustinsson LE, et al. Electrical stimulation versus coronary artery bypass surgery in severe angina pectoris: the ESBY study. Circulation. 1998;97:1157-1163. 3. Gibbons RJ, Abrams J, Chatterjee K. et al; American College of Cardiology; American Heart Association Task Force on practice guidelines (Committee on the Management of Patients With Chronic Stable Angina). ACC/AHA 2002 guideline update for the management of patients with chronic stable angina—summary article: a report of the American College of Cardiology/American Heart Association Task Force on practice guidelines (Committee on the Management of Patients with Chronic Stable Angina). J Am Coll Cardiol. 2003;41:159-168. 4. Lanza GA, Sestito A, Sandric S, et al. Spinal cord stimulation in patients with refractory anginal pain and normal coronary arteries. Ital Heart J. 2001;2:25-30. 5. Jessurun GA, Hautvast RW, Tio RA, DeJongste MJ. Electrical neuromodulation improves myocardial perfusion and ameliorates refractory angina pectoris in patients with syndrome X: fad or future? Eur J Pain. 2003;7:507-512. 6. Lanza GA, Sestito A, Sgueglia GA, et al. Effect of spinal cord stimulation on spontaneous and stress-induced angina and ’ischemia-like’ ST-segment depression in patients with cardiac syndrome X. Eur Heart J. 2005;26:983-989.

The American Journal of Medicine (2007) 120, e19

LETTER Opioid Agreement or Patient-Centric Action Plan? To the Editor: As an academically educated, board certified pain management physician, my opinions surrounding use of opioid agreements differ from those recently stated by Arnold et al.1 Chronic disease states require a commitment from both physician and patient in order to actualize therapy and optimize outcomes. Arnold and colleagues are correct in stating that use of “opioid agreements” may lack empirical evidence, but “chronic disease action plans,” which model opioid agreements, have been widely utilized and studied in patient-centric chronic disease management and demonstrated significant advantages in improved patient care.2-4 Ellwood launched the concept of patient-centric care in1988 and has continued efforts to underscore the importance of collaborative patient-physician disease therapy.5 In 2003, Johns Hopkins and American Healthways convened to further define the patient-physician relationship. One recommendation included use of verbal or written agreements to improve communication and treatment outcomes.6 An opioid agreement is a simple and reproducible tool designed to define the responsibilities of both the clinician and patient and provide the patient with detailed risks and benefits of the prescribed therapy. It offers the opportunity to document critical concerns, such as personal issues, safe medication storage and use, the role of adjuvant testing, therapeutic alternatives, expected outcomes, and the consequences of nonadherence.7 I offer the following response in regard to the authors’ perceived ethical dilemma surrounding the use of opioid agreements, as stated “From an ethical standpoint, the discriminatory implementation of opioid contracts may potentially violate patient’s rights to fair and equal treatment.” Not all patients suffering with chronic pain are candidates for opioid therapy, but all chronic disease patients may benefit from a written treatment plan. Patients who do require opioid therapy, especially those with a prior history of abuse or addiction, are entitled to a comprehensive therapeutic management plan and are expected to adhere.8 The authors seem to understand the essence of an opioid agreement, as stated “Apart from their function of securing

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.

informed consent, opioid contracts may serve to organize and provide rational order to the problem of treating chronic, nonmalignant pain.” Order and documentation are both essential components, especially within the current medico-legal environment. But the authors caution the reader “Clinicians should exercise caution when deciding if and how they use this potentially promising but clearly problematic intervention.” This statement is bold and unmerited and only serves to remind us of how far we still have to go in understanding the complexities of chronic pain management. Clinicians should definitely incorporate a written treatment agreement for all pain patients. To quote Dr. Donald Berwick, Clinical Professor of Pediatrics and Health Care Policy at Harvard Medical School, President and CEO of the Institute of Healthcare Improvement, and an early patientcentric care advocate, “Nothing about me, without me.” Jeffrey A. Gudin, MD Englewood Hospital and Medical Center Englewood, NJ

doi:10.1016/j.amjmed.2006.07.042

References 1. Arnold RM, Han PK, Seltzer D. Opioid contracts in chronic nonmalignant pain management: objectives and uncertainties. Am J Med. 2006; 119:292-296. 2. Williams DM. Considerations in the long-term management of asthma in ambulatory patients. Am J Health Syst Pharm. 2006;63(10 Suppl 3):S14-S21. 3. Handley M, MacGregor K, Schillinger D, Sharifi C, Wong S, Bondenheimer T. Using action plans to help primary care patients adopt healthy behaviors: a descriptive study. J Am Board Fam Med. 2006;19:224-331. 4. Turnock AC, Walters EH, Walters JA, Wood-Baker R. Action plans for chronic obstructive pulmonary disease. Cochrane Database Syst Rev. 2005;(4):CD005074. 5. Ellwood PM. Shattuck lecture— outcomes management. A technology of patient experience. 1988. Arch Pathol Lab Med. 1997;121:11371144. 6. American Healthways. Johns Hopkins University. Defining the PatientPhysician Relationship for the 21st Century. 3rd Annual Disease Management Outcomes Summit. 2003. Available at: http://www.patientphysician.com/docs/PatientPhysician.pdf. Accessed June 9, 2006. 7. Burchman SL, Pagel PS. Implementation of a formal treatment agreement for outpatient management of chronic nonmalignant pain with opioid analgesics. J Pain Symptom Manage. 1995;10:556-563. 8. Resnick DB. The patient’s duty to adhere to prescribed treatment: an ethical analysis. J Med Philos. 2005;30:167-188.

The American Journal of Medicine (2007) 120, e21

LETTER Opioid-induced Androgen Deficiency Discussion in Opioid Contracts To the Editor: The excellent article by Arnold et al1 prompted us to suggest an addition to the text of most opioid contracts. It is our practice to improve the usefulness of these contracts to both patients and practitioners by including in them a discussion of opioid-induced androgen deficiency (OPIAD). Strong inhibition of androgen production quickly follows the onset of sustained-action opioid use, whether the opioids are administered orally,2-5 transdermally,4 or intrathecally.3,5-8 These low androgen levels result in classic symptoms of hypogonadism in a majority of opioid-consuming men, with most of them exhibiting various combinations of fatigue, depression, hot flashes, night sweats, diminished libido, erectile dysfunction, and diminished sexual arousal and satisfaction. Low androgen levels also contribute to physical changes, including osteoporosis, anemia, and diminished muscle mass. Similar changes occur in opioid-consuming premenopausal women, most of whom demonstrate amenorrhea or anovulatory menstrual cycles soon after beginning sustained-action opioid use.6,8-10 Most premenopausal and postmenopausal women also demonstrate greatly diminished libido6,8,10 at this time. Signs and symptoms of hypogonadism typically improve in men with OPIAD during replacement testosterone therapy4 and in the few women who have received hormone replacement therapy for opioid-related symptoms of hormone deficiency.6,10 In my private practice of general internal medicine, the discussion of OPIAD in our opioid contracts includes the potential usefulness of replacement hormonal therapy and potential complications of this therapy. We have found this information to be of help in establishing the patient-physician partnership described and recommended by Arnold et al,1 which clearly is required for optimal chronic use of opioids. As part of this information exchange, we obtain sex-hormone levels in all outpatients about to begin chronic

0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.

sustained-action opioid therapy and again with the development of symptoms of hypogonadism. Our contract includes discussion of testicular atrophy as a frequent complication of opioid therapy, which commonly progresses more rapidly during testosterone replacement therapy for OPIAD.4 Although this atrophy may become profound, patients are generally reassured to know that this complication is not accompanied by changes in penile size. It is our current practice to routinely document testicular size by using a Prader orchidometer before beginning chronic opioid therapy. Harry W. Daniell, MD Department of Family Practice University of California Davis Medical School, Redding

doi:10.1016/j.amjmed.2006.05.027

References 1. Arnold R, Han P, Seltzer D. Opioid contracts in chronic nonmalignant pain management: objectives and uncertainties. Am J Med. 2006;119: 292-296. 2. Daniell H. Hypogonadism in men consuming sustained-action oral opioids. J Pain. 2002;3:377-384. 3. Rajagopal A, Vassilopoulou-Sellin R, Palmer JL, et al. Symptomatic hypogonadism in male survivors of cancer with chronic exposure to opioids. Cancer. 2004;100:851-858. 4. Daniell H, Lentz R, Mazer N. Open-label pilot study of testosterone patch therapy in men with opioid-induced androgen deficiency. J Pain. 2006;7:200-210. 5. Roberts LJ, Finch PM, Pullan PT, et al. Sex hormone suppression by intrathecal opioids: a prospective study. Clin J Pain. 2002;18:144-148. 6. Abs R, Verhelst J, Maeyaert J, et al. Endocrine consequences of long-term intrathecal administration of opioids. J Clin Endocrinol Metab. 2000:85;2215-2222. 7. Paice J, Penn R, Ryan W. Altered sexual function and decreased testosterone in patients receiving intraspinal opioids. J Pain Symptom Manage. 1994;9:126-131. 8. Paice JA, Penn RD. Amenorrhea associated with intraspinal morphine. J Pain Symptom Manage. 1995;10:582-583. 9. Schmittner J, Schroeder JR, Epstein DH, Preston KL. Menstrual cycle length during methadone maintenance. Addiction. 2005;100:829-836. 10. Daniell HW. Opioid endocrinopathy in women consuming prescribed sustained-action opioids for control of non-malignant pain. J Pain. In press.

APM Perspectives

APM Perspectives The Association of Professors of Medicine (APM) is the national organization of departments of internal medicine at the US medical schools and numerous affiliated teaching hospitals as represented by chairs and appointed leaders. As the official sponsor of The American Journal of Medicine, the association invites authors to publish commentaries on issues concerning academic internal medicine. For the latest information about departments of internal medicine, please visit APM’s website at www.im.org/APM.

Clinical Simulation: Importance to the Internal Medicine Educational Mission Paul E. Ogden, MD, Lauren S. Cobbs, MD, Martha R. Howell, MEd, Stephen J. B. Sibbitt, MD, Donald J. DiPette, MD Department of Internal Medicine, Scott & White Memorial Hospital and Clinic, Scott, Sherwood and Brindley Foundation, The Texas A&M University System Health Science Center College of Medicine, Temple.

Medical education in the United States is constantly evolving. However, some aspects have remained unchanged since the Flexner Report of 1910. One aspect that hasn’t changed is the strong apprenticeship component of clinical education during which medical students and residents train in academically oriented hospitals supervised by faculty physicians.1 Recently, this apprenticeship model for medical education began undergoing an extreme makeover.2,3 Beginning in 2001, the Accreditation Council for Graduate Medical Education (ACGME) Outcome Project required residency programs to provide education and competency assessment of each resident in 6 core competencies: Patient Care, Medical Knowledge, Communication, Professionalism, Practice-Based Learning and Improvement, and SystemsBased Practice.2 The Liaison Committee for Medical Education has similar expectations for the evaluation of undergraduate medical education.4 Medical knowledge is expanding at a much faster rate than it can be learned. One solution for medical educators and trainees to minimize potential knowledge gaps has been to lengthen the duration of training and to increase specialization. The Liaison Committee for Medical Education and ACGME also have responded by increasing the curricular mandates to medical schools and residency training programs. Requests for reprints should be addressed to Paul E. Ogden, MD, Department of Internal Medicine, Scott & White Memorial Hospital and Clinic, 2401 South 31st Street, Temple, TX 76508. E-mail address: [email protected]

However, increased curricular mandates and longer, more specialized training have natural limitations. Patient safety is now a national priority, and patient safety education will need to follow this movement. Following the publication of the Institute of Medicine report, To Err is Human, the federal government and health care providers made a significant financial commitment to creating a safer system of care.5 Many of the preventable medical errors involve faulty systems of care, lack of standardization in practice, poor communication, and lack of teamwork among health care providers.6 High-fidelity simulation (full-body mannequin robots that breathe, talk, blink, and respond physiologically “like a real person”) provides an opportunity to compress the learning time in medicine through a high quality experiential learning environment.7 The curriculum can be taught in a reproducible and emotionally engaging manner that significantly enhances understanding and retention, and reduces gaps in knowledge and skills.8 A high-fidelity simulation program can reproducibly teach and assess most, if not all, of the ACGME competencies under direct observation. Patient safety issues can be simulated and practiced so that mastery of skills occurs in a safe environment. Mistakes can occur without “real life” consequences. High-fidelity patient simulation has been described as a method to not only teach, but evaluate while replicating a variety of clinical encounters.9-12 A comprehensive simulation program should become as natural to medical education as teaching

0002-9343/$ -see front matter © 2007 The Association of Professors of Medicine. All rights reserved. doi:10.1016/j.amjmed.2007.06.017

Ogden et al

Residency Education, Patient Safety, High Fidelity Patient Simulation

821

since the 1960s but were expensive and too sophisticated for the users.14,15 Beginning in the late 1980s, anesthesia mannequins were developed that could generate physiologic changes to mimic critical clinical events. In the last 10 years, mannequin simulators have improved significantly in levels of reTHE CASE FOR SIMULATION alism.14 Simulation manneSystematically designed simulaquins can reproduce physitions are not new. They have been PERSPECTIVES VIEWPOINTS ologic changes associated employed since ancient times in ● Simulation can reproducibly teach and with unstable cardiac and those pursuits where training and respiratory conditions, and assess ACGME competencies. testing in the real world have been have advanced airway featoo dangerous (eg, war games), too ● Medical education has been slow to tures that can simulate diffiexpensive (eg, aviation), or simply adopt simulation technology, a mainstay cult airway scenarios. unfeasible (eg, space exploraSeveral partial-task for training in several other high-risk tion).13 Educational simulations are trainers are available that occupations. analogous to the play of any speallow practice of technical cies. By allowing imitation, the par- ● Discuss the basic components of and skills in a safe environticipant can explore, make miscosts and challenges to incorporating ment before attempting takes, and incorporate corrective simulation technology. these procedures on pafeedback as a guide for future tients. This technology has action.8 developed to the point that Comprehensive patient simulacredentialing for some tion programs are quickly emerging in many US medprocedures, such as carotid stenting, now requires demical schools.14 Departments of internal medicine have onstration of competency on a simulator before perforlagged behind other clinical departments (such as anmance on a patient.19 esthesia and emergency medicine) in the use of highFull body mannequins were designed initially for fidelity simulation for medical education. However, to anesthesia training and historically are best suited for date, no medical specialty has utilized the full potential dealing with anesthesia and advanced cardiac life supof high-fidelity simulation as a teaching tool. port scenarios.8 Because internal medicine training Medical simulation—as a teaching and assessment heavily involves diagnostic skills, simulation scenarios tool— offers several advantages over traditional patient are best done with additional layers, such as obtaining encounters. First, events can be scripted so that specific a history using an embedded microphone in the mancurricular objectives are experienced at the same time nequin and using standardized patient actors to play the by all of the trainees. This is possible for both common 12 roles of family members, nurses, or other patients. and rare scenarios. Second, events can be observed in Scenarios can be scripted to reproduce complex sitreal time and allowed to unfold. Mistakes can occur uations, including ethical dilemmas, communication without endangering patient safety.11,12,15-18 Third, challenges, risk management issues, team dynamics, trainees can be placed under significant pressure where and cultural challenges. Simulation also is ideal for ethical decision-making, cultural awareness, and comteam training, including rapid response and code munication skills must be employed. Finally, clinical teams. It has been the authors’ experience that wellsimulation allows the opportunity for practice and repscripted simulation can provide trainees with cues etition until a skill is mastered.11,17,18 In other words, in and consequences similar to those encountered in the this setting, the competency of the trainee can be obhospital. served and documented. Omissions in the curriculum High-fidelity simulation is best described as a stratcan be identified. Remediation and practice is possible, until the desired outcome for each trainee can be egy, not a technology; it is a teaching tool that can be obtained. used to mirror, anticipate, or amplify real clinical situations for the purpose of standardizing the curriculum and compressing the timeframe during which experiHUMAN PATIENT SIMULATION—HISTORY, ences are normally encountered. These experiences are designed to promote reflective learning by uncovering STATUS, AND POSSIBILITIES and challenging the trainee’s otherwise hidden percepSimulation training has been a mainstay for training in tions. Learners can try different strategies to solve several high-risk occupations, including aviation, nuproblems and practice skills. They also can rehearse the clear power, and the military.14 However, medical edmanagement of serious but infrequent events under ucation has been slow to adopt simulation technology. Realistic human patient simulators have been available controlled, predictable times and places.8 rounds and morning report. Because internal medicine has traditionally led the education of medical students and educational reform, internal medicine should also lead in simulation education.

822

The American Journal of Medicine, Vol 120, No 9, September 2007

COMPONENTS OF A SIMULATION PROGRAM Simulation requires 3 basic components: a trained faculty member to develop and deliver the curriculum, simulation equipment, and a space that provides adequate realism for the suspension of disbelief. The faculty member’s role is to deliver the curriculum using simulation as the tool. Gordon et al have shown that the same case can be used with any level of trainee as long as the curricular objectives are clear.7 For example, a case of an inferior myocardial infarction can be used with first- and second-year students to demonstrate cardiac physiology, pathology, or pharmacology. The same case used with third-year students can demonstrate diagnosis, electrocardiogram interpretation, and initial management. For a postgraduate year-1 resident, the case can focus on management and cardiac care unit (CCU) care; with upper level residents, the emphasis can shift to advanced cardiac care, management of shock, or code team training.7 Different simulation equipment can be used depending upon the goals of the session. Sessions can be designed primarily for education, training (practice), evaluation, or combinations of the above. Examples of equipment include partial task trainers (such as a lumbar puncture trainer or central line trainer), computer software that runs patient scenarios, virtual reality, and various computerized mannequins. The higher fidelity mannequins, such as SimMan (Laerdal Medical Corporation USA, Gatesville, Tex) and METI (Medical Educational Technologies Inc., Sarasota, Fla), can reproduce vital signs and some physiologic changes. These mannequins have voice capability, allowing oral history-taking, but are limited in terms of physical examination capabilities. Standardized actors can add layers of complexity to scenarios, especially if the goals of the session are to evaluate communication, professionalism, or ethics (Table 1). Several types of educational scenarios can be built into a simulation program. Sessions can be designed to teach and practice procedures. Code team training, rapid response team training, and critical event scenarios, such as mass casualty or weapons of mass destruction scenarios, can use simulation to teach and practice skills that cannot be taught safely during normal rotations. Scott & White Memorial Hospital recently used simulation to identify response problems and team co-

Table 1

Common Simulation Modalities

Partial task trainers (ie, Lumbar Puncture Trainer) Standardized patients Full body task trainers (ie, Resusci Anne) High fidelity mannequins (ie, SimMan, METI) Virtual reality Computer software (ie, MicroSim)

Table 2

Common Types of Scenarios

Task/procedure training (eg, central line insertion, lumbar punctures) Team training (eg, ACLS, code team training) Critical event scenarios (eg, difficult airway, WMDs, mass casualties) Behavioral/cognitive skills training (eg, communication, professionalism, ethics) Competency assessment ACLS ⫽ advanced cardiac life support; WMD ⫽ weapons of mass destruction.

ordination problems for code teams, working in the context of moving into a new hospital. The hospital simulated unannounced codes with mannequin simulators to uncover logistical problems and improve team coordination. With some imagination, almost anything in the curriculum can be simulated. Simulation is an excellent tool for teaching cultural competence, professionalism, ethics, and communication. Scripted scenarios, using standardized family members, friends, and hospital personnel, have been used to challenge the resident’s ethical training and reasoning. These scenarios include unclear end-of-life decisions (patient “codes” with ambiguous code status), conflicts surrounding the transfusion of a patient (patient is unable to communicate but does not have written “do not transfuse” orders), and disclosure of a death to family members. Risk management can be extremely helpful in identifying common at-risk events (Table 2). Issenberg et al published a systematic review of the essential features of simulation that lead to effective learning, finding that feedback was the most important feature.20 The authors’ experience also has strongly reinforced the concept that all simulation sessions should be followed immediately by a structured debriefing session. During the debriefing session, the learner has an opportunity to reflect on his or her actions and beliefs. The resultant “reflective learning” mirrors the type of learning that occurs with difficult cases in the hospital. An active learner in a stressful situation must make decisions and can immediately see the consequences of those decisions.20 The authors’ motto (adopted from the Harvard Institute for Medical Simulation) is that “simulation is just an excuse for debriefing.”8 Other key features identified by Issenberg et al include repetitive practice, curriculum integration, variation in degree of difficulty, the use of multiple learning strategies, ability to capture clinical variation, and defined outcome measures.20

CHALLENGES AND COST Although simulation is gaining acceptance in medical education, several challenges must be addressed before

Ogden et al

Residency Education, Patient Safety, High Fidelity Patient Simulation

a simulation program can become an educational standard. The major challenges include space, time, faculty buy-in, and money. Many institutions cannot afford to build and maintain a simulation center. Simulation centers are a significant capital expenditure and simulation technology also can be expensive to obtain and maintain. It has been the authors’ experience that declining clinical revenue has made educational space more difficult to obtain. Scott & White Memorial Hospital took advantage of a different strategy, partnering with a community college, Temple College. Temple College constructed a $3 million simulation center within a new $8 million Health Sciences building. Scott & White health system was willing to provide equipment and some of the mannequins, and to share in the annual expense of the center. Temple College uses the center for their surgical technology, respiratory technology, emergency medical technology, and nursing programs. The center also is used for training of residents from Darnall Army Hospital at Fort Hood, Texas, and individuals associated with regional emergency services. Another significant challenge to simulation is obtaining time for both the trainees and faculty. Residents are essential to patient care and their time has been restricted by ACGME duty hour regulations.2 Residency directors may be reluctant to move residents away from the inpatient setting into a place that is educationally unfamiliar, especially when the residents are critical to the medical care delivered by the institution. The authors found faculty, student, resident, and fellow buy-in to be a slow process. Therefore, a schedule was created so that there would be minimal disruption to daily clinical and academic responsibilities. The faculty buy-in process for medical student learning included overcoming resistance to changing the curriculum and taking time away from other activities. The authors incorporated the simulation course during the outpatient block of the third-year internal medicine clerkship. The curriculum committee did not immediately endorse the time commitment or the curriculum, but allowed a pilot to be conducted. The students strongly endorsed the simulation experience and now the student curriculum committee enthusiastically supports the use of simulation. Simulation training is now a significant part of the “educational culture” of Scott & White Memorial Hospital, Texas A&M University HSC College of Medicine. The key players in changing attitudes and perceptions were the students and residents who immediately realized the value of simulation training. The simulation program also is part of the recruitment strategy, so the students and residents expect it once they arrive.

823

The program requires several faculty members to be successful. Within the department of medicine, approximately 1 full-time equivalent (FTE) is dedicated to the program, with time commitments being divided between 4 regular clinical faculty members and a fulltime medical educator.

THE FUTURE The airline industry adopted simulation as a major teaching and evaluation method more than 60 years ago in response to a poor safety record, scrutiny from the Federal Aviation Administration, and incentive to reduce the number of crashes during pilot training. Simulation is now the standard for training and evaluation within the industry. All pilots are required to pass a checkout ride in the simulator before flying the actual plane. They also must recertify at regular intervals and must train and certify when flying any new airplane. As a consequence, the airline industry has an enviable safety record. Simulation should become as natural for physicians as it is for pilots. Most procedures can be simulated by novices before they are attempted on patients. The fidelity of partial task trainers and virtual reality trainers is rapidly improving and is realistic enough for practice and basic skill acquisition.21 Although simulation is not universally practical today, all procedural skills should eventually be simulated and practiced up to a basic level of competency before they are ever attempted on patients. In the near future it should become unthinkable for trainees to practice and acquire basic skills on patients. As the patient safety movement matures, simulation will most likely play a significant role in ongoing training and recertification. For example, the airline industry shares all data about crashes and near misses. All of those events are carefully evaluated along with the basic procedures and systems that may have contributed to the crash. If the problem was procedural, the standard procedures are then changed. The whole process then is brought to the simulator and becomes a part of normal training and recertification (personal communication with J.E. Buttry, retired American Airlines Captain of 16 years, on January 12, 2006). The same process could be used with medical simulation. Risk management cases can be evaluated in the same way as airline crashes. Subsequent procedural changes could be simulated and become part of the certification process. Overcoming cultural obstacles, such as sharing risk management data and standardization of hospital procedures, will be daunting, but the health care industry cannot continue to tolerate mistakes and preventable deaths as the cost of an imperfect system. An important use of clinical simulation as a teaching tool is its ability to define and create a “critical inci-

824

The American Journal of Medicine, Vol 120, No 9, September 2007

dent” in which a medical learner is made aware of or recognizes an area of clinical deficiency and therefore becomes motivated to learn and incorporate new knowledge.22 Simulation should be a core element in the education of the physician throughout the educational continuum. In the earlier stages of medical school, simulation can be used to introduce physiological and pathological concepts and to practice clinical skills. Later in medical school, simulation can be used to provide standardization to the curriculum, to teach technical skills, and to observe noncognitive skills. Residents and faculty can benefit greatly from simulation for all the reasons listed above, as well as to practice skills and react to serious events that are usually rare in clinical presentation. The same scenarios can be used throughout the educational continuum by simply varying the level of complexity and by adding confounding factors. Clinical simulation also allows for complex areas of previously acquired medical information, skill sets, and professional behaviors to be integrated. Discernment and retention can occur more efficiently than with conventional teaching methods, particularly when it comes to the areas of clinical ethics, professionalism, and communication skills.7 Finally, simulation offers a standardized, reproducible method for delivery of competency education. Simulation also allows for direct observation of residents and students in stressful situations. Internal medicine should take a leading role in this revolution. Since the days of Sir William Osler, internal medicine has been a leader in the clinical education of medical students and residents. Simulation should be no exception. This educational method will become the standard for competency assessment in the near future and be as routine for the continuum of physician education as it is for aviation and other high stakes professions. With departments of internal medicine leading the way, internal medicine training programs can train a better, safer, and more competent physician for the 21st century.

References 1. Cooke M, Irby DM, Sullivan W, Ludmerer KM. American medical education 100 years after the Flexner Report. N Engl J Med. 2006;355:1339-1344. 2. Accreditation Council for Graduate Medical Education. ACGME outcome project. Available at: http://www.acgme.org/outcome. Accessed January 10, 2007.

3. Gisondi MA, Smith-Coggins RS, Harter PM, et al. Assessment of resident professionalism using high fidelity simulation of ethical dilemmas. Acad Emerg Med. 2004;11:931-937. 4. Liaison Committee on Medical Education (LCME). Revised standards adopted February 6-8, 2007. Supplement ED-1A. Available at: http://www.lcme.org. Accessed January 10, 2007. 5. Institute of Medicine. To Err is Human: Building a Safer Health System. Washington, DC: National Academy Press; 1999. 6. Leape L, Epstein AM. A series on patient safety. N Engl J Med. 2006;347:1272-1273. 7. Gordon JA, Oriol NE, Cooper JB. Bringing good teaching cases “to life”: a simulator-based medical education service. Acad Med. 2004;79:23-27. 8. Instructor training. Presented Institute for Medical Simulation as a Teaching Tool at Harvard Medical School. October 4-6, 2006. 9. McLaughlin SA, Doezema D, Sklar DP. Human simulation in emergency medicine training: a model curriculum. Acad Emerg Med. 2002;9:1310-1318. 10. McMahon G, Monaghan C, Falchuk K, et al. A simulator-based curriculum to promote comparative and reflective analysis in an internal medicine clerkship. Acad Med. 2005;80:84-89. 11. Kneebone R, Scott W, Darzi A, Horrocks M. Simulation and clinical practice: strengthening the relationship. Med Educ. 2004; 10:1095-1102. 12. Vozenilek J, Huff J, Rezneck M, Gordon J. See one, do one, teach one: advanced technology in medical education. Acad Emerg Med. 2004;11:1149-1154. 13. McGuire CH. Simulation: its essential nature and characteristics. In: Tekian A, McGuire CH, McGaghie WC (eds). Innovative Simulations for Assessing Professional Competence: From Paper and Pencil to Virtual Reality. Chicago, IL: University of Illinois at Chicago, 1999:3. 14. Cooper JB, Taqueti VR. A brief history of the development of mannequin simulators for clinical and education training. Qual Saf Health Care. 2004;13(Suppl 1):i11-i18. 15. Dalley P, Robinson B, Weller J, Caldwell C. The use of high fidelity human patient simulation and the introduction of new anesthesia delivery systems. Anesth Analg. 2004;6:1737-1741. 16. Fried M, Satava R, Weghorst S, et al. Identifying and reducing errors with surgical simulation. Qual Saf Health Care. 2004; 13(Suppl 1):i19-i26. 17. Patow C. Improving patient safety through simulation technology. AHIP Cover. 2004;3:74-77. 18. Friedrich MJ. Practice makes perfect: risk-free training with patient simulators. JAMA. 2002;288:2808-2812. 19. Gallagher AG, Cates CU. Approval of virtual reality training for carotid stenting. What this means for procedural-based medicine. JAMA. 2004;292:3024-3026. 20. Issenberg SB, McGaghie WC, Petrusa ER, et al. Features and uses of high fidelity medical simulations that lead to effective learning: a BEME systematic review. Med Teach. 2005;27: 10-28. 21. Reznick RK, MacRae H. Teaching surgical skills— changes in the wind. N Engl J Med. 2006;355:2664-2669. 22. Morell VW, Sharp PC, Crandall SJ. Creating student awareness to improve cultural competence: creating the critical incident. Med Teach. 2002;24:532-534.