The International Primary Pulmonary Hypertension Study (IPPHS) The International

Primary

Pulmonary

Hypertension

P

rimary pulmonary hypertension (PPH) is an illdefined and very rare disease. Depending on the definition used, its annual incidence has been estimated to be around 1 per 200,000 to 1 per 1,000,000 inhabitants in countries like France or the United S t a t e s . By definition, PPH has served as a general term to designate pulmonary hypertension which could not be directly attributed to an obvious or sufficient cause. From an epidemiologic viewpoint, this approach is, however, barely satisfactory since PPH is thus relegated to the status of a symptom without a cause, similar to the so-called "essential hypertension." Probably because of its rarity, PPH has remained largely unexplored as to its risk factors. Throughout the years, the potential role of factors such as age, gender, pregnancy, systemic diseases, diet, drugs, and others have been prop o s e d . Also, genetic and immunologic factors are suspected to play a role in the development of the disease. None of these factors has actually been quantitatively assessed. Most of them are not even truly confirmed to play a role. 13

343

1415

The International Primary Pulmonary Hypertension Study (IPPHS) is an international effort established in order to compensate for some of the above-mentioned epidemiologic deficiencies. Its main objective is to quantitatively assess the role of several known or suspected risk factors in the development of PPH. More specifically, the role of age, gender, obesity, medical and surgical history, obstetric history, comorbidity, smoking, alcohol, caffeine, oral contraceptives, selected classes of drugs, as well as some environmental factors, will be assessed. Due to the rarity of the disease and the complexity of the problem at hand, the design and implementation of the study raised several methodologic issues. This paper intends to present an overview of the design of the study and the options chosen to solve some of these methodologic difficulties. As well, the limitations of the study will be presented. OVERVIEW OF THE STUDY

The rarity of the disease excluded a prospective

Study

Group"

field cohort study. The difficulties associated with the diagnosis of PPH would have made a "recordlinkage" database study hazardous. Consequently, a field case-control study was the design of choice. Cases are being recruited in five countries: France, the United Kingdom, Belgium, the Netherlands, and Switzerland (in chronological order of participation into the study) from specialized centers: Pneumology, Cardiology, and Transplantation units. Four control subjects are recruited for each case. These are "physician-based controls." The intent is to recruit 100 validated cases and 400 properly matched control subjects for the case control analysis. Some of the methodologic difficulties encountered also militate in favor of the recruitment of a number of extra cases to be used for verification purposes. All cases and control subjects are interviewed by specially trained interviewers using a specially designed questionnaire. The study was launched in October 1992 in France and in early 1993 in the other countries. So far, 150 centers have agreed to participate in the study, and approximately 65 cases have been recruited, of which two thirds will certainly be kept for the case-control analysis. CASES

Definition of Cases Cases are defined as patients with pulmonary hypertension without a sufficient cause for this symptom. The list of the causes thought to be sufficient acTable 1—Excluded Conditions Congenital abnormalities o f t h e lungs, thorax, o r diaphragm Congenital o r a c q u i r e d valvular o r myocardial disease Pulmonary thromboembolism Obstructive lung disease Interstitial lung disease P u l m o n a r y artery o r p u l m o n a r y valve stenosis P u l m o n a r y venous hypertension C e n t r a l hypoventilation with hypoxemia and hypercapnia Parasitic diseases affecting the lungs Sickle cell a n e m i a

" D r s . Lucien Abenhaim, C h a i r m a n ; Yola M o r i d e ; Stuart Rich; Anicet Chaslerie; F r a n c o i s Brenot; Tim H i g e n b o t t a m ; Celia Oakley; X a v i e r Kurz; Maurits D e m e d t s ; E m i e l F . M . W o u t e r s ; G e r a l d S i m o i m e a u ; Cornelis A. W a g e n v o o r t ; and Michel Aubier.

AIOS (Definite collagen vascular disease) Active liver disease

C H E S T / 105 / 2 / FEBRUARY, 1994 / Supplement

Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21690/ on 01/27/2017

37S

cording to literature and consensus which leads to the exclusion of cases is presented in Table 1. Suspected risk factors or conditions associated with pulmonary hypertension that are not considered so far as a sufficient cause are not considered as reasons for excluding the cases. These include conditions such as elevated ANA titers, suspected or not clinically-overt systemic diseases, portal hypertension, liver diseases, HIV infection, exposure to drugs, and others, along the lines defined by the ACCP Consensus Statement. 16

Case

Ascertainment

Cases?

The requirement to validate all cases and to be able to interview the patients precluded a retrospective recruitment. Thus, only cases diagnosed after January 1, 1992, were considered. Given that the study was launched after this date, the design allowed for the inclusion of so-called "prevalent" cases (diagnosed between January 1, 1992, and August 30, 1992) and "incident" cases (diagnosed after the launching of the study). Al-

Table 2—Case Ascertainment • C a s e s are r e p o r t e d by participating c e n t e r s to local r e s e a r c h teams in each country. • A chest physician o r cardiologist in e a c h local r e s e a r c h t e a m (country) insures t h e s c r e e n i n g o f t h e case o v e r t h e p h o n e with t h e reporting clinician. • W h e n cases are p r e a c c e p t e d , this medical specialist visits t h e reporting c e n t e r for reviewing t h e c a s e and for extraction o f data from t h e medical charts. • An international panel o f experts reviews all cases (including xrays and scans) and classifies cases into ( 1 ) appropriate cases, ( 2 ) possibly appropriate c a s e s , and ( 3 ) r e j e c t e d cases.

38S

CHOICE OF CONTROL SUBJECTS

The source population for the cases is one of the first criteria for die selection of the type of control subjects. The source population for the cases in the IPPHS can only be defined secondarily (secondary base) to the recruitment. There are important discrepancies in the criteria used for diagnosis between centers, so it cannot be assumed that all cases will eventually be diagnosed as PPH. Consequently, the probability of being diagnosed as PPH seems to be related to the chance of being referred to a given center. The source population for each case reported by this center should theoretically be the catchment population of this center. However, because patients with PPH were first suspected by a referring physician, it is not the usual catchment population of the reporting center that should be considered. Rather, it is the catchment population for referral of PPH. Such a population is extremely difficult to define because the referring physicians could be very different for each case, and the population that they serve could also be very heterogeneous. The feasibility study has shown that cases could be referred by pulmonary or heart specialists, by any medical specialists practicing in public or private institutions, or even by general practitioners. 17

Each reported case is ascertained through a three-step process: (1) A number of investigations are deemed necessary for the diagnosis to be confirmed; the list of these "mandatory" investigations can be found in Table 2. It is verified, over the phone, with the reporting physician, that these investigations have been conducted. (2) A pulmonary or cardiology specialist extracts the data from the medical charts of each patient; there is one such specially trained physician in each countiy. (3) All cases are subsequently reviewed by an international panel of experts. This panel reviews the medical extraction form supplemented by the lung perfusion scan, and the chest x-ray films. Each patient is classified into the three following categories: (a) definitively appropriate for the study (accepted cases); (b) possibly appropriate (possible cases); and (c) not appropriate for the study (rejected). Cases can be rejected because a sufficient cause of PPH is found in one of the steps described above or because, while actually being "real" PPH cases, they missed some important information, and a consensus cannot be reached. Prevalent or Incident

though the two sets of cases are not very distant from a calendar viewpoint, it is necessary to consider them, to a certain extent, separately because some patients may have died in the interval. If more severe cases had a different history of exposure or different risk factors than the others, that is, if one risk factor for the disease also had an influence on the probability of survival, considering both sets together could lead to a potential bias.

It was, however, not felt that community control subjects would be the optimal choice because the probability of being diagnosed with PPH could depend on ones access to the medical system. At least the delay between the onset of prodromic symptoms—such as dyspnea— and the diagnosis could be much shorter in those patients regularly followed by physicians as opposed to those not followed. Because one of the suspected risk factors investigated is a prescription drug, control subjects should have the same probability as the patients to have this drug prescribed to them, should they need it. It was necessary to control for the prescription of drags by the physician, at least for a sample of cases. This was difficult to obtain with community control subjects. Another criterion to take into consideration is to ensure that the distribution of exposure in the control group is similar to the distribution in the source population (study base). (Hospital control subjects were not considered.) Because the IPPHS investigates simultaneously a large number of suspected risk factors, ensuring that all selected hospital diagnoses were independent of the probability of exposure to any of these factors seemed an impossible task. T h e International Primary Pulmonary Hypertension Study (tPPHS

Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21690/ on 01/27/2017

Group)

Physician-based control subjects have been proposed as an interesting alternative to both hospital control and community control subjects. This type of control subject is more representative of the general population than most hospital control subjects while controlling for the referral bias that could possibly jeopardize the validity of community control subjects. However, this type of control subject is not without limitations as discussed below. The first limitation of physician-based control subjects is the feasibility of the complexity of the logistics. In effect, there are numerous steps as follows: (1) identification of the patient's general practitioner or of a properly matched neighboring physician; (2) list of patients seen by the physician during 1 week from which are selected four properly matched control subjects per case; (3) authorization for contacting and interviewing these patients; (4) interviewing these patients in their environment (because they are matched to the case on the area of residence, this can require important travelling of the interviewers in each country). At each step, consent is required from all parties. Measures are taken to ensure confidentiality of information gathered. This process was tested and found feasible but is costly with respect to time, resources, and manpower. Another limitation is the possible overmatching for some characteristics, in particular for drags, since it is believed that physician habits might have little variation in this matter. A simulation of the effect of this possible overmatching has found it to be minimal in this case. An important concern with information obtained by interviews is the possibility of recall bias. Because patients with PPH are suffering from a serious life-threatening disease, it is very likely that recall of their past exposure to drugs would differ from that of relatively healthy individuals. One way to minimize this source of bias would be to use patients with another severe condition. However, for the reasons described above, it was not possible to use hospital control subjects. Furthermore, it would have been difficult to find hospitalized conditions in the gender and age groups considered (mainly females 20 to 50 years) in which hospital incidence would be high enough to produce four properly matched control subjects per case. Recall bias is a serious concern and therefore important to address in the interview procedure. It is important to note that, to a certain extent, the recall bias could work in both directions. On one hand, patients with PPH could be more likely to recall suspected exposure because they would be searching for a possible reason to their health status. On the other hand, it was found that in some instances, subjects tended to deny exposure to certain factors, such as licit or illicit drugs, RISK FACTORS INVESTIGATED

There are few epidemiologic data available on the risk

factors for the disease. Most of the knowledge in this regard emerges from case series. The role of factors such as age and gender still remains controversial because of the heterogeneity between studies in the definition of PPH. The study addresses the main risk factors usually investigated in epidemiology, such as age, gender, familial history, medical history, surgical history, obstetric history, smoking, caffeine consumption, and a number of lifestyle habits (Table 3). The primum movens for the study was the allegation regarding the potential role of anorectic agents in the development of the disease. Given that a number of other drugs has been reported to be potentially associated with PPH, a thorough ascertainment of exposure to drugs is necessary. Actually, all therapeutic classes of drugs are being investigated, some of the basis of a priori hypotheses as to their potential association with PPH, and others either as markers of morbidity conditions or in order to ensure some blindness vis a vis the incriminated drugs. Classes of drugs which have been suspected to play a role in the development of PPH are, beside anorectic agents, the following: thyroid extracts and amphetaminelike drugs, oral contraceptives, and tryptophan (a diet supplement). There is also some biologic plausibility in the potential role of drugs interfering with the metabolism of serotonin. All of these factors are assessed by interviewing cases and control subjects. Medical history and exposure to drugs are assessed through spontaneous reporting by the subject as well as being elicited by a closed questionnaire. A special questionnaire, structured around personal life events, has been designed. For drugs, visual displays of packages or pictures are also used. For variables considered more confidential or sensitive (HIV infection, sexual history, illicit drug use), an anonymous questionnaire to be filled out by the subject is used. Nonmedically trained interviewers were recruited in order to minimize the biases potentially associated with solicitation of responses. They are blind as to the objectives of the study. It is also intended to subsequently collect blood samples for future genetic and physiologic studies. Table 3—Risk Factors Investigated Questionnaire I (Interviews) • Age, gender, s o c i o e c o n o m i c status, o c c u p a t i o n • Smoking, caffeine, alcohol consumption • Medical history • O b s t e t r i c history • D r u g s (including oral contraceptives, diuretics, beta-blockers, s y m p a t h o m i m e t i c s , anorexigens, amphetamine-like, thyroid extracts, psychotropic drugs, others) • A m p h e t a m i n e s , c o c a i n e , IV drugs, o t h e r "illicit" drugs

C H E S T / 105 / 2 / FEBRUARY, 1994 / Supplement

Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21690/ on 01/27/2017

39S

CONTROL OF BIASES

Sample Distortion

Biases

These biases, often referred to as "selection biases," can occur as a result of a distortion in the inclusion of cases or of control subjects. Distortion can occur in the diagnosis as well as in the reporting of cases. The diagnosis of PPH is a difficult one, being one of exclusion. It is facilitated by the presence of a suspected risk factor, such as one of the drugs listed above. Worst, cases of secondary pulmonary hypertension could wrongly be diagnosed as primary only because of the presence of one suspected risk factor for PPH. The first type of bias would result in a number of PPH cases remaining undiagnosed because they would not be exposed to a risk factor. On the basis of present knowledge, such a bias is considered to be of small magnitude in Western countries because PPH occurs in an age group where a serious, often fatal disease with severe dyspnea is unlikely to remain uninvestigated. Nevertheless, there is still room for some misclassification. Of more concern is the second type of bias. The potential role of anorectic agents has received important publicity is some countries. Cardiologists and pulmonary physicians are well aware of the aminorex experience. Thus, the discovery of exposure to an anorectic agent in patients presenting with pulmonary hypertension could lead to a premature termination of clinical and paraclinical investigations. The same type of bias could occur with other factors: it has been suggested that a recent obstetric history, or HIV infection, could also be viewed by some clinicians as strong indications of the primary cause of an observed pulmonary hypertension. To control for this bias, only patients who have undergone all mandatory diagnostic procedures listed in Table 2 are considered for the study. On the basis of these procedures, the international panel of experts reviews all cases blindly as to their exposure to suspected risk factors. Another more subtle sample distortion bias could occur if the diagnosis was only accelerated in patients exposed to a suspected risk factor. This would result in survival bias whereby exposed patients would appear to survive longer after the diagnosis. Also, those patients would be more likely to be exposed in a given time-window. This bias cannot be controlled for at the stage of the study design but will be considered at the analysis stage. Although the recruitment of cases intends to be as exhaustive as possible in each country, it is not expected that an objective of exhaustivity of recruitment could be realistically reached due to the rarity of the disease. Thus, there is room for biases in the reporting of cases. Symmetrical biases in opposite directions are envisioned. Cases in which patients were exposed to suspected risk factors would be more likely to be reported because clinicians would consider them as more interesting for the 40S

study. Patients exposed to some risk factors would not be reported because such cases would be considered as secondary. The same risk factors could be considered as "known" or only as "suspected" depending on the center and/or clinician involved. To control for these biases, whenever a case is reported by a center, the list of discharge diagnoses of all the patients seen in this center in the preceding 12 months will be examined in order to identify the ones that could potentially be missed. Also, a sample of centers that reported no cases will be examined. Depletion of susceptible patients could occur if patients exposed to certain risk factors would withdraw from exposure when prodromic symptoms or signs would develop and if such a withdrawal would lead to a reversibility of the disease process. This phenomenon would bias the results towards the null by underestimating the number of exposed cases. One mitigation strategy vis a vis this bias is to collect as much information as possible on the previous history of both patients and control subjects, which is done here for most variables at hand. However, there is little ground to consider that this phenomenon could account in the case of PPH, which is considered as a progressive disease by many. A number of the issues related to the sample distortion biases resulting from the selection of control subjects have been addressed above. Confounders Several sources of confounding were considered, some of which have been detailed above. Among other sources, particular attention was given to confounding by indication. Confounding by indication could occur if drugs were used in individuals who were more susceptible to develop PPH because of some underlying conditions. As an example, if obesity was a risk factor for PPH, it could be confounded with the utilization of anorectic agents (or the reverse). That is why an important emphasis has been put on the design stage in collecting information on a number of medical variables that will be tested as potential confounders. Finally, another bias to consider could be the so-called "protopathic bias." This bias could occur if an exposure to a given agent would occur as a result of the development of prodromic symptoms or signs. This exposure would, therefore, be wrongly considered as the risk factor for the disease. For instance, certain drugs could be used in the presence of dyspnea in its early stages, and those drugs could be wrongly considered as responsible for the development of the full-blown syndrome of PPH. To address this issue, an important effort is made to determine as accurately as possible the date of the beginning of symptoms (mainly dyspnea). This date is considered as the index date for the study, and exposure before this date for patients and control subjects is compared. T h e International Primary Pulmonary Hypertension Study (IPPHS

Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21690/ on 01/27/2017

Group)

Ethical Considerations Because the study is supported by a pharmaceutical company that could have some interest in the results, an independent International Scientific Board has been established. Close links are also kept with the regulatory authorities of each country involved. If needed, these authorities have access to the research files if there was such a need. Efforts are made to preserve the condifentiality of the information collected. The consent of patients and subjects enrolled in the study is sought at each stage. CONCLUSION

The study of the risk factors for a disease as rare as PPH is not an easy task. There is a certain amount of discrepancies between centers as to the definition of the criteria for the disease. However, such an effort is considered as largely justified by the poor prognosis for most patients not responding to treatment, for which prevention is the only available strategy. The knowledge of risk factors for the disease should help establish such preventive strategies. COMPOSITION OF THE IPPHS GROUP

Members of the International Scientific Board Profs. Lucien Abenhaim; Michel Auhier; Bernard Begaud; Jacques Benichou; Maurits Demedts; Tim Higenbottam; Celia Oakley; Stuart Rich; Gerald Simonneau; Cornelis A. Wagenvoort; Emiel F.M. Wouters Members of Local Research Teams France and Switzerland: Drs. Frangois Brenot, Anicet Chaslerie, Claudine Peiffer The Netherlands and Belgium: Drs. Hans Petri; Xavier Kurz; Robert Naeije; Denise Walckiers United Kingdom: Drs. Mark Ryan, Neal Uren Scientific Consultants Drs.William Dab; David Langleben; Bruno Strieker; Emmanuel Weitzenblum Members of the International Reviewing Panel Profs. Stuart Rich, Chairman; and I^ewis Rubin; Drs. David Langleben, Michael D. McGoon

A C K N O W L E D G E M E N T S : W e want to acknowledge the Institut d e R e c h e r c h e s International Servier for its financial support to the study and t h e G o u v e r n e m e n t d e Belgique. This study could not b e possible without the participation o f m o r e than 1 5 0 clinical c e n t e r s in the five countries involved ( F r a n c e , the U n i t e d Kingdom, the Netherlands, B e l g i u m , and Switzerland).

REFERENCES 1 Rich S. P r i m a r y p u l m o n a r y h y p e r t e n s i o n . P r o g C a r d i o v a s c Dis 1988; 31:205-38 2 Blin P, A b e n h a i m L . A survey o f p r i m a r y p u l m o n a r y hypertension diagnosis in t h e Aquitaine region o f F r a n c e : preliminary r e p o r t ( s u b m i t t e d for publication) 3 R i c h S, D a n t z k e r D R , Ayres S M , Bergofsky E H , B r u n d a g e B H , D e t r e K M , et al. P r i m a r y p u l m o n a r y h y p e r t e n s i o n : a national p r o s p e c t i v e study. Ann I n t e r n M e d 1 9 8 7 ; 1 0 7 : 2 1 6 - 2 3 4 B r e n o t F, H e r v e B , Rain B , S i m o n n e a u G. H y p e r t e n s i o n arterielle p u l m o n a i r e primitive ( D o n n e e s d e la l i t t e r a t u r e c t e x p e r i e n c e p e r s o n n e l l e sur 1 2 5 cas e n 1 0 a n s ) . 1 9 9 1 ; L a R e v u e du Praticien 1991; 17:1560-67 5 Kleiger

R,

Boxer

M,

Ingham.

Harrison

D.

Pulmonary

h y p e r t e n s i o n in patients using oral c o n t r a c e p t i v e s : a report o f six cases. C h e s t 1 9 7 6 ; 6 9 : 1 4 3 - 4 7 6 O a k l e y C, S o m e r v i l l e J . O r a l c o n t r a c e p t i v e s a n d p r o g r e s s i v e p u l m o n a r y vascular disease. L a n c e t 1 9 6 8 ; 1 : 1 8 9 0 - 9 3 7 M l c z o c h J . D r u g and dietary i n d u c e d p u l m o n a r y hypertension. In: W e i r K, R e e v e s JT, eds. P u l m o n a r y hypertension. Mt. Kisco, NY: F u t u r a Publishing, 1 9 8 4 ; 3 4 1 - 5 8 8 F o l l a t h F, B u r k a r t F, S c h w e i z e r W. D r u g - i n d u c e d p u l m o n a r y hypertension? B M J 1 9 7 1 ; 1:265-66 9 H r u b a n Z. P u l m o n a r y c h a n g e s i n d u c e d by a m p h o p h i i c drugs. E n v i r o n Health P e r s p e c t 1 9 7 6 ; 1 6 : 1 1 1 - 1 8 1 0 W h i t e JP, W a r d M J . D r u g - i n d u c e d adverse p u l m o n a r y reactions. A d v e r s e D r u g R e a c t A c u t e Poisoning R e v 1 9 8 5 ; 4 : 1 8 3 - 2 1 1 11 C u r t n e r HP. P u l m o n a r y h y p e r t e n s i o n , " p l e x o g e n i c p u l m o n a r y a r t e r i o p a t h y " and t h e a p p e t i t e d e p r e s s a n t drug aminorex: post or p r o p t e r ? Bull E u r o p Physiopath R e s p i r 1 9 7 9 ; 1 5 : 8 9 7 - 9 2 3 1 2 A t a n a s s o f P, W e i s s B , S c h m i d E , T o r n i c M .

Pulmonary

h y p e r t e n s i o n and d e x f e n f l u r a m i n e . L a n c e t 1 9 9 2 ; 3 3 9 : 4 3 6 13 R o c h e N, L a b r u n e S, B r a u n e J - M , H u c h o n G J . P u l m o n a r y h y p e r t e n s i o n and d e x f e n f l u r a m i n e . L a n c e t 1 9 9 2 ; 3 3 9 : 4 3 6 - 3 7 1 4 B a r s t R G , F l a s t e r E R , M e n o n A, F o t i n o M , M o r s e J H . E v i d e n c e f o r the association o f u n e x p l a i n e d p u l m o n a r y hypertension in c h i l d r e n with t h e m a j o r histocompatibility c o m p l e x . Circulation 1992; 85:249-58 1 5 M e l m o n K, B r a u n w a l d E . Familial p u l m o n a r y hypertension. N Engl J M e d 1963; 2 6 9 : 7 7 0 1 6 Rubin L J . P r i m a r y p u l m o n a r y h y p e r t e n s i o n : A C C P c o n s e n s u s statement. Chest 1993; 1 0 4 : 2 3 6 - 5 0 17 W a c h o l d e r S, M c L a u g h l i n G D , S i l v e r m a n D T , M a n d e r G S . Selection o f c o n t r o l s in c a s e - c o n t r o l studies: I. Principles. Am J Epidemiol 1992; 1 3 5 : 1 0 1 9 - 2 8

C H E S T / 105 / 2 / FEBRUARY, 1994 / Supplement

Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21690/ on 01/27/2017

41S

PULMONARY MEDICINE AT YOUR FINGERTIPS u

D

A CCP-SEEK Keep abreast of recent advances in the treatment of pulmonary diseases with ACCP-SEEK

Vol. 3: Pulmonary

A

r

i

Volume 3: Pulmonary Medicine

• Assess your current knowledge in pulmonary medicine. • Earn up to 50 hours of Category 1

Medicine, a self-assessment program

CME Credit at your convenience

for physicians seeking a comprehen-

upon completion of Volume 3.

sive overview of pulmonary medicine. ACCP-SEEK: Pulmonary Medicine

problem-solving skills relevant to

• Keep your practice up-to-date. • Learn from renowned experts offering the most recent advances in diagnosis and management of pulmonary diseases.

COLLEGE

pulmonary medicine. • Confidential scoring of answers by American College Testing (ACT), Iowa City. • Individual and group score reports sent to you by ACT. • A printed rationale addressing each test item.

OF

CHEST PHYSICIANS.

• 200 multiple-choice questions

• Prepare for Board Certification/

study program.

1-800-343-ACCP

Volume 3 includes: that test recall, interpretation and

with this comprehensive self-

ORDER YOUR COPY TODAY!

AMERICAN

helps you: Recertification Examinations

D

Accreditation Statement: The American College of Chest Physicians is accredited by the Accreditation Council for Continuing Medical Education (ACCME) to sponsor continuing medical education. CME Credit: The American College of Chest Physicians designates this CME activity for up to 50 hours in Category 1 of the Physician's Recognition Award of the American Medical Association. Forfurther information, contact Registration and Product Information American College of Chest Physicians 3300 Dundee Road Northbrook, Illinois 60062 1-800-343-ACCP (Fax: 708-498-5460)

Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21690/ on 01/27/2017