Title: A systematic review of the prevalence of signs of impending death and symptoms in the

Running Title: Symptoms in the Last Weeks 1 Title: A systematic review of the prevalence of signs of impending death and symptoms in the last two we...
2 downloads 2 Views 1MB Size
Running Title: Symptoms in the Last Weeks


Title: A systematic review of the prevalence of signs of impending death and symptoms in the last two weeks of life

Authors: Karen A Kehl, PhD, RN, ACHPN University of Wisconsin – Madison School of Nursing 600 Highland Ave K6/354 CSC Madison, WI 53792 Phone: 608-265-2190 Fax: 608-263-5458 Email: [email protected]

Jennifer A Kowalkowski, MS, MPH, RN University of Wisconsin – Madison School of Nursing

Running Title: Symptoms in the Last Weeks


Abstract To optimally manage patient care, knowledge of the prevalence of signs of impending death and common symptoms in the last days is needed. Two reviewers independently conducted searches of PubMed, CINAHL, PsychINFO and the Web of Knowledge from 1996-2011. Peer reviewed studies of adults that included contemporaneous documentation of signs and symptoms were included. Reviewers independently extracted data. Twelve articles, representing 2416 patients in multiple settings were selected. Of the 43 unique symptoms, those with the highest prevalence were: dyspnea (56.7%), pain (52.4%), respiratory secretions/death rattle (51.4%), and confusion (50.1%). Overall prevalence may be useful in anticipating symptoms in the final days and in preparing families for signs of impending death.

Keywords: end-of-life care, symptoms, signs of impending death, dying, symptom management, symptom prevalence

Running Title: Symptoms in the Last Weeks


Introduction Understanding the signs of impending death and symptoms that are commonly experienced in the last weeks of life is crucial for directing appropriate care for patients and those who support them. A number of studies have identified the signs and symptoms typically noted in the weeks before death, as well their prevalence 1-20, but methodological limitations threaten the generalizability of these findings. Best evidence for certain health states, like the last weeks of life, cues health care professionals in their assessment and treatment of signs and symptoms of patients in their care. Therefore, more reliable data on the frequency with which signs and symptoms occur in the last weeks of life increases the likelihood that signs or symptoms will be noted and consequently treated. This knowledge will improve information for health care professionals who provide care for dying individuals in a variety of settings, as well as information for informal caregivers who provide care for dying individuals in the home. The last two weeks of life, commonly referred as the terminal period or “actively” or “imminently” dying presents numerous challenges for both professional and informal caregivers. During this period, new symptoms may develop, previously managed symptoms may change, and signs of impending death often occur. As death approaches, symptom management is complicated by the patient’s declining ability to communicate. When informal caregivers are the primary care providers, such as in home hospice, symptom management becomes more complicated. Families frequently report feeling unprepared for the physical changes in the patient in the last week of life 21. There is a need for evidence-based information about what signs of approaching death and symptoms to expect 22 to improve assessment and management of common symptoms in the terminal period for both professional and informal caregivers.

Running Title: Symptoms in the Last Weeks


A number of studies that characterize the dying process have been published over the last 20 years. Individual studies have focused primarily on identifying signs and symptoms of impending death, or validation of data collection tools. Much of the existing research on signs and symptoms in the last weeks of life has focused on signs and symptoms associated a single diagnosis, like cancer 1,12,20,23, obstructive pulmonary disease 5,18 , end-stage renal failure 9, or heart failure 2,8,11,14, while others have focused on specific symptoms like dyspnea or delirium 13,16

. In other studies, identification of signs and symptoms were reported as a secondary goal


. A number of reviews concerning signs and symptoms in those near the end of their lives

have been published in the last ten years 22-26. Four of these were systematic reviews 22,24,25 26 and the fifth was an integrative review 23. A variety of methods for data collection were employed in the individual studies included in these reviews, such as audits of medical records, concurrent data collection for patients prospectively identified for study inclusion, and interviews with family members or health care professionals after the death in which they recalled the events prior to the death. Plonk and Arnold also included expert opinion data from textbooks and other published sources 22. One review 26 excluded data of symptoms that were assessed by proxy. The review only included patients with a cancer diagnosis that were not receing curative treatment. Variability in data collection methods likely contributes to the wide ranges of sign and symptom prevalence reported across reviews. In the most recent of these reviews, published in 2011, the reviewers examined symptoms and associated outcomes in patient with advanced cancer. The review found that while large number of symptoms (56) were evaluated, different studies examined different symptoms 24 . In the 22 studies analyzed in that review, only 12 provided information on prevalence. In each of these reviews, there was great variability in the

Running Title: Symptoms in the Last Weeks


prevalence of signs and symptoms. For example, reported prevalence of dyspnea ranged from 10-95% 25, pain from 8-96% 4,25 , respiratory secretions from 25-92% 25, confusion from 6-93% 25

, and nausea/vomiting from 6-68% 25. Many of these previous studies and reviews have included retrospective recall of

symptoms by providers or family members. Studies have reported that symptom recall is poor when data about signs and symptoms are collected following the death 27-29. To date, no systematic review or meta-analysis has been published reporting the prevalence of signs and symptoms experienced in the last days of life solely from data reported by studies in which data was collected contemporaneously, or as the signs and symptoms occurred and that includes multiple diagnoses.

Objectives The purposes of this systematic review were to identify commonly occurring signs of impending death and symptoms that occur in the last two weeks of life and to estimate their overall prevalence.

Methods Literature Search Methods A review protocol was developed by the researchers in advance of the literature search to guide the inclusion and analysis of studies. Only published articles were sought. Each reviewer independently searched designated databases (Table 1). Additional articles were identified through ancestral searching and citation tracking 30, as well as a hand search of the contents

Running Title: Symptoms in the Last Weeks


pages of the Journal of Hospice and Palliative Nursing. The last search was conducted on May 17, 2012. Groups of keywords and subject headings were combined to search each database (Table 1). Limits were set to humans, adults, and publication dates 01/01/1996-12/31/2011. Similar strategies were employed for each database searched. No limits to publication language or patient diagnosis were imposed. Articles retrieved from ancestral searches were included, regardless of year of publication, if they met the inclusion criteria. The searches were repeated after initial analysis of the findings using the same databases, keywords, and limits except the publication dates were expanded to 01/01/1996-05/17/2012 to assure inclusion of the most current literature.

Inclusion/exclusion criteria Peer reviewed empirical studies that reported the prevalence of physical signs or symptoms in the last two weeks of life were included in the review. Only those studies where data collection occurred contemporaneously with the presentation of signs and symptoms were included. To improve the validity of findings and reduce the risk of bias, studies that used retrospective recall to collect prevalence of signs or symptoms were excluded from review. Articles were excluded if they assessed symptoms by proxy or gave only data on symptom intensity (without specifying prevalence). In addition, those studies focusing on signs or symptoms related to the withdrawal of life support were excluded since typical signs and symptoms might be suppressed by advanced life support or the process of withdrawal of life support. Articles without clear information to determine adherence to the inclusion criteria were also excluded.

Running Title: Symptoms in the Last Weeks


Screening and data extraction Articles were managed in EndNote X6 (Thomson Reuters, New York, NY). Each reviewer conducted an independent search of the selected databases and initially screened titles and abstracts from search results to eliminate articles that were not research or did not include data concerning signs and/or symptoms in the last two weeks of life. Relevant abstracts were independently screened for inclusion. If there was not enough information to determine whether to retain the article in the review, or if an article appeared to meet the inclusion criteria, the full text of the article was obtained. Each reviewer nominated articles for inclusion in the review independently based on the inclusion and exclusion criteria. After review for inclusion and exclusion criteria was completed, the reviewers completed a STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) checklist and eliminated studies that did not have clear information on the source of the data and methods of assessment or that did not report prevalence or numbers of patients with each sign or symptom 31. After assessment for quality, the reviewers jointly discussed the remaining articles and the final sample was determined through consensus. The reviewers independently extracted data from the selected studies using a reviewerdeveloped data extraction form. The form was based on previously published systematic reviews, including a review of symptoms in lung cancer patients, and modified for the purposes of this study 32. Extraction data were compared and disagreements resolved through re-evaluation of the study and consensus between the reviewers. Information extracted from each study included: 1) sample size, 2) setting, 3) study design, 4) methods of data collection (including rationale for symptoms assessed and data collection tools), 5) diagnoses of study participants, 6) assessment

Running Title: Symptoms in the Last Weeks


period (time until death), 7) number of symptoms evaluated, 8) prevalence for each symptom, and 9) main study findings.

Analysis The primary outcome measures were the signs and symptoms documented and the overall prevalence of those signs and symptoms across studies, both weighted and unweighted. Studies were categorized as “restricted” (r) if the investigators identified specific signs and symptoms a priori for examination in the study, and “unrestricted” (u) if not. The number of patients reporting each of the signs and symptoms was attained by summing the number of subjects from the individual studies which evaluated that signs or symptoms . For some studies 6,33-36 the number of patients reporting each signs and symptoms was calculated based on the percent reported and the sample size. These numbers were calculated independently by both reviewers and then compared for accuracy. Both unweighted (u) and weighted (w) prevalences were calculated for each sign and symptom included in the review. The weighted prevalence was determined by calculating the mean for each sign or symptom across studies: ([(sum of # patients with signs or symptoms across studies)/(# patients assessed across studies)]*100). The unweighted prevalence was determined by computing the average percent for each sign and symptom across studies: ((sum of % with signs or symptoms in each study)/(# studies assessing signs or symptoms)). Weighted prevalence calculations are likely more representative of the population prevalence given that more emphasis is placed on those results from studies with large sample sizes. Measures of central tendency were calculated for the unweighted prevalence. Data were

Running Title: Symptoms in the Last Weeks


synthesized and displayed using tables and Forrest plots. The prevalence of symptoms in the restricted studies was compared to the prevalence in the unrestricted studies using an independent samples t-test to examine the potential for selection bias. SPSS 19.0 was used for all calculations.

Results Study Selection The two reviewers performed independent screening of each study for relevance and eligibility as shown in Figure 1. Based on firm application of inclusion criteria, a total of 12 articles were retained for final analysis in the systematic review 6,33-43.

Descriptions of the Studies The studies included in the review represent data collected in five countries; the United States, Canada, Spain, Hong Kong, and Japan. A variety of settings were examined, including inpatient medical centers 33,35,40,43, and palliative care units 38,41,42, as well as long-term care facilities 6, outpatient clinics 34, and home hospice 36,37,39. Four studies 37,38,41,42 included only subjects with a diagnosis of cancer and an additional six 6,33,35,39,40,43 included both cancer and non-cancer diagnoses. Two studies evaluated signs and symptoms in specific non-cancer populations, end-stage renal disease 34 and heart failure 36. The studies examined a range of 24 hours to two weeks preceding the death, with five studies examining the last seven days. The characteristics of each study are summarized in Table 2.

Running Title: Symptoms in the Last Weeks


Two methods for data collection were reported in the studies reviewed. Eleven of the studies employed data extraction from the medical record, including provider notes, and one study obtained data by interviewing patients during their last seven days about the presence of symptoms at the time of the interview 37. Different tools and approaches were used in each study to assess signs and symptoms and record data. Six studies, labeled “restricted” 6,35,37,41-43 only assessed for the presence of predetermined signs or symptoms. The other six studies, labeled “unrestricted” 33,34,36,38-40 did not determine in advance the signs or symptoms assessed for the study. The number of signs or symptoms included in each study ranged from one to 27 (Table 2). Although there was overlap in the signs and symptoms examined across studies, no two studies reported data on the same group of signs and symptoms. The quality assessment revealed a range in the number of STROBE criteria addressed in each included study. None of the studies included all of the criteria from the STROBE checklist (Appendix 1). Four of the 12 studies addressed the greatest number, including 18 of the 29 components and subcomponents the STROBE checklist, while one included study included only eight.

Prevalence of Signs and Symptoms In total, 62 signs and symptoms in the final two weeks of life were identified across. Some signs and symptoms were named differently across studies, but described in the same manner. These signs and symptoms were collapsed with the broadest term or the term most commonly used in the literature retained as the final symptom descriptor. Table 2 shows the terminology used by the individual study investigators. In all cases, the reviewers attempted to

Running Title: Symptoms in the Last Weeks


preserve the integrity of the symptom descriptors from the original study. The terms collapsed into a single symptom category are listed below with the retained term in bold: Abdominal swelling and ascites Agitation, delirium and restlessness Confusion, confusional state Fatigue or lethargy Respiratory secretions, death rattle, sputum, and pulmonary congestion Skin integrity problems, smelly wounds, sores, and decubitus ulcers Sleep problems, sleep disturbance and insomnia Weakness and asthenia Confusion was not included with agitation, delirium and restlessness since the descriptions of the conditions varied and some studies 35,39,40,43 reported both as separate symptoms. Nausea/vomiting was reported in four studies as only nausea 33-35,38 in one study as only vomiting 37 and in two studies 39,40 as nausea/vomiting. In total, nineteen signs and symptoms were collapsed leaving a total of 43 unique signs and symptoms identified across studies (Table 3). There was wide variation in prevalence of symptoms ranging from 8.6% to 55.7% (Table 3 & Figure 2). For symptoms reported in multiple studies, the most commonly reported symptoms were dyspnea (62.1% w, 56.1% u), weakness (54.4% w, 23.9% u), respiratory secretions (53.3% w, 51.4% u), and pain (47.2% w, 52.8% u). In contrast, of the signs and symptoms reported in multiple studies, pruritus (1.2%w, 1.7% u) was the least prevalent, followed by depression (8.3% w, 10.6% u), diarrhea (8.7% w, 9.6% u), sleep problems/insomnia (9.0% w, 11.5% u) and urinary problems (10.1 % w, 10.9% u).

Running Title: Symptoms in the Last Weeks


There were notable differences between the unweighted and the weighted prevalence for numerous signs and symptoms (Table 4). For constipation, dyspnea, fever, and weakness, the weighted prevalence was greater than the unweighted prevalence. For agitation, anxiety, confusion, depression, fatigue, nausea, pain, and skin integrity problems, the unweighted prevalence was greater than the weighted prevalence. In addition, for agitation and nausea there is one study with a prevalence that greatly differs from the other studies reporting that symptom. Table 4 describes the variation in prevalence of signs and symptoms. There was a significant difference in the prevalence of three symptoms between the restricted studies (r), in which specific symptoms were assessed for and/or extracted from the medical record and unrestricted studies (u), which did not specify symptoms a priori, as determined by an independent samples t test. The prevalence of anxiety (8.98% u, 35.75% r; 2 tailed sig, -3.133(4) = 0.035), depression (2.53% u, 38.60% r; 2-tailed sig, -21.959(2)= 0.002) and sleep problems other than increased sleeping (6.85% u, 28.40% r; 2 tailed sig, -3.965(3)= 0.029) were significantly lower in the unrestricted studies than in the restricted studies. Edema was noted only in unrestricted studies. All other signs and symptoms were either reported in a single study, or no significant difference was found between the restricted and unrestricted prevalence.

Discussion The findings from this review are congruent with previously reported ranges of prevalence of signs and symptoms in the last weeks of life, but provide additional knowledge of overall prevalence by the calculation of a weighted or overall prevalence for each sign or symptom. The prevalence of dyspnea has previously been reported as 10-95% 25; the weighted

Running Title: Symptoms in the Last Weeks


prevalence of 62.1% and the range of 25.0-78.0% reported here is consistent with that range and the findings from the SUPPORT study in which more than 50% of those who died in the hospital experienced dyspnea 10. Similarly, the ranges of prevalence for pain (42.2%, range 23.0-99.0%; previous reports 8.0% 4 to 96.0% 25), respiratory secretions (53.3%, range 35.0-92.0%; previous reports 25% 44 to 92% 42), confusion (37.6%, range 18.2-86.0%; previous reports 6% to 93% 25) and nausea/vomiting (19.4%, range 8.4-71.0%; previous reports 6% to 68% 25) that are reported in this review are similar to findings from previously published studies and reviews 4,22,24-26,42,44 . Given how broad the prevalence ranges are from previous studies, those ranges would be inclusive to most findings as can be seen in the large unweighted confidence intervals (Figure 2). For a number of signs and symptoms (confusion, dyspnea, fatigue, and diarrhea) this review describes a narrower range than previous reviews. There were no symptoms that had a higher prevalence in this review than reported in prior reviews. However, four signs and symptoms; Agitation/delirium/restlessness (20.8%, range 5.8-51%; previous reports 28% to 83% 45), anxiety (10.8 %, range 1.4-45.5%; previous reports 8-75% 25), depression (8.3%, range 0.9-38.6%; previous reports 3-82% 25) and sleep problems/insomnia (9.0%, range 3.2-28.4%; previous reports 9-74% 25 ) were somewhat lower than previously reported ranges. Several factors may contribute to the inconsistency between the findings in this review as compare to previous reviews. Data on signs and symptoms obtained from family caregivers may be inflated or minimized based on their level of distress or discomfort in response to the patient’s condition, particularly in situations where the patient is unable to communicate. In addition, certain signs and symptoms may have been reported inaccurately based on poor or inadequate definitions.

Running Title: Symptoms in the Last Weeks


Examining the unweighted mean symptom prevalence with confidence intervals (Figure 2) shows that cough, depression, myoclonus, pruritus, sleep problems, urinary problems and weakness have confidence intervals that cross zero, making the results difficult to interpret. Some signs and symptoms, notably anorexia, fatigue, fever and respiratory secretions have wide confidence intervals. This may be related to the heterogeneity of the sample and the range of assessment techniques used. For these signs and symptoms the mean prevalence calculation is not as robust as it is for agitation, dyspnea, nausea, pain and decubitus, where the confidence intervals are relatively narrow. Six studies included in this review 6,34,35,38,41,42 were also analyzed in the Plonk and Arnold review 22, which may account for some of the consistency in findings. However, the findings from this review are also similar to those from the Solano, Gomes and Higginson review25, despite there being no overlap of studies included between the reviews. The lower prevalences reported in this review may be due to the narrow inclusion criteria of this review. The review by Teunissen, et al 26 also excluded proxy assessment of signs and symptoms but included only cancer patients. There were some pooled preferences from the Teunissen, et al review that were very similar to what was found in this review, especially regarding constipation, hemorrhage, nausea/vomiting, pain, and skin problems. The pooled preferences reported by Teunissen, et al for other signs and symptoms were either higher (anorexia, anxiety, depression, fatigue and weakness) or lower (confusion, cough, diarrhea, dry mouth, dysphagia, dyspnea, urinary problems) than the overall weighted prevalence reported here. The similarities and differences between the Teunissen, et al review and this review elicits a number of possible reasons for the difference most obviously the range of diagnoses included. Limitations

Running Title: Symptoms in the Last Weeks


Only one of the included studies was published in the last five years. The search was conducted to include all studies from 1996 to May, 2012. Despite the pace of change in palliative and end-of-life care and the considerable research concerning symptoms, few studies have provided concurrent assessment of symptoms in the final two weeks of life. We screened some studies that may have included data concerning symptoms, but it was not presented in a manner that allowed us to extract prevalence of symptoms specifically during the last two weeks of life. Other studies examined symptoms over time, provided information about symptom severity, medications given to manage symptoms or symptoms as prognostic indicators but not appropriate data to determine symptom prevalence in the last two weeks of life. As new technology provides us with new ways to collect data on symptoms, we may see more reports of symptom prevalence and trajectories. There was a wide range of prevalence of the signs and symptoms, sample sizes, and data collection methods reported across these studies. These factors likely contributed to the differences between the weighted and unweighted signs and symptoms prevalence reported. Additionally, the inclusion of data from a variety of settings and diagnoses may have contributed to the wide ranges of signs and symptoms observed. Documentation practices may play a significant role in the reliability of these findings. For example, the number of signs or symptoms included in the studies reviewed ranged from one to 27, and most of the studies (11 of 12) utilized medical record extraction for data collection. Discrepancies between how often a sign or symptom occurred and how often it was documented in the medical record are possible given that documentation typically represents a single encounter with a patient rather than findings from continuous assessments or from the outpatient setting. Those signs and symptoms that are not featured in documentation templates of data

Running Title: Symptoms in the Last Weeks


collection instruments, or in the standardized assessment forms may be underreported or not reported at all. Conversely, those signs and symptoms that are either featured prominently or are required by the institution may be over reported. In addition, studies have confirmed that completeness of documentation of signs and symptoms for patients who are dying diminishes as the patient approaches death 4 , so it is possible that as the patients’ death neared, documentation of symptoms may be less complete. The study by Bailey, et al analyzed in this review reports an intervention to improve the quality of care for patients dying in an acute care inpatient setting. The study reports documented prevalence of symptoms before and after the intervention. The increase in documentation of signs and symptoms in the days prior to death can be attributed largely to the effects the intervention had in increasing awareness of signs and symptoms in patients who were actively dying 33. There was little difference in prevalence of most symptoms between studies that restricted the symptoms assessed and those that did not. The symptoms with a significant difference between restricted and unrestricted prevalence, anxiety, depression and sleep problems, may be documented less frequently if there is not an assessment instrument that cues the clinician. It is also possible that the data extracted from the reviewed studies are biased or incomplete based on a priori restrictions of signs or symptoms which did not include these particular symptoms. Indeed the restricted studies described an average of eight signs and symptoms, where the unrestricted studies described an average of 13.7 signs and symptoms. Another limitation relates to how symptoms were categorized. For example, bowel problems might include constipation and/or diarrhea. When a specific symptom was identified, it was retained as a category for this review. It was clear during the review process that some signs and symptoms were named differently, but had similar descriptions. In these situations,

Running Title: Symptoms in the Last Weeks


categories that included multiple terms were created to provide a larger sample for each symptom group for the purposes of analysis. The reviewers acknowledge that different combinations of terms may have yielded different prevalences, however every attempt was made to preserve the integrity of the symptom descriptors from the original study. The heterogeneity of this sample was a limitation that may have contributed to the wide range of prevalence that was described. We thought that since many who provide care to those at the end of life see patients with a variety of diagnoses, that including studies with different diagnoses would provide a broad picture of the prevalence of common signs and symptoms. We included a broad range of settings for the same reason. The information may also help identify symptoms that are common across settings or diagnoses that should be a focus for future research. Finally, inherent to the quality of any systematic review is the quality of the studies included. Over the 15 years spanned by the articles in this review, expectations and standards for peer-reviewed publications have changed, especially with the application of reporting standards such as the CONSORT 46 or STROBE


standards. While some of the studies included do not

meet every item on the STROBE checklist, there was sufficient detail in the article to assure that the study met the inclusion criteria.

Future research

There is clearly a need for additional, high quality, prospective studies of the prevalence of signs and symptoms of impending death. There is a particular need for information about prevalence in the home hospice setting, where informal caregivers are at the bedside and

Running Title: Symptoms in the Last Weeks


medications and treatments need to be anticipated to optimize care. Prospective studies examining signs and symptoms from the patient and/or informal caregiver perspective could add to our understanding of the experience of the patient and caregiver in the final weeks of life. We made the decision to focus this review on physical signs and symptoms, yet there are many other conditions that can cause emotional, spiritual and existential suffering at the end of life. A number of other reviewers have addressed these issues 47-50. Future research examining the interactions between the various causes of suffering for dying individuals could add to our understanding and suggest novel ways to manage suffering in the last weeks of life. There were a number of questions about signs of impending death and symptoms in the last days that we were not able to determine from the data in these studies. Future research should be designed to allow comparison of signs and symptoms in cancer and specific noncancer diagnoses. Information about symptom clusters in the final days and symptom trajectories would further add to clinicians’ ability to optimally manage signs and symptoms that occur as death approaches. In particular, examination of the trajectory of symptoms in the weeks prior to death might provide valuable information for better determination of prognosis in the final days, which could affect medical decision-making. Further advances in clinicians’ abilities to improve care, enhance quality of life, and reduce suffering in the final phase of life should be based on accurate data that is collected through expert assessment and documentation by those who provide direct patient care. The weighted prevalences reported in this review provide the best knowledge to date of the likelihood that a particular patient may experience one of the signs or symptoms addressed in these studies. This knowledge can guide assessment of symptoms, planning for symptom management and preparation of family members for the changes in the final weeks of life.

Running Title: Symptoms in the Last Weeks


References 1.

Addington-Hall J, McCarthy M. Dying from cancer: results of a national population-based investigation. Palliative medicine. Oct 1995;9(4):295-305.


Anderson H, Ward C, Eardley A, et al. The concerns of patients under palliative care and a heart failure clinic are not being met. Palliative medicine. Jul 2001;15(4):279-286.


Butters E, Higginson I. Two HIV/AIDS community support teams: patient characteristics, problems at referral and during the last 6 weeks of life. AIDS Care. 1995;7(5):593-603.


Ellershaw JE, Smith C, Overill S, Walker SE, Aldridge J. Care of the dying: Setting standards for symptom control in the last 48 hours of life. Journal of pain and symptom management. 2001;21(1):12-17.


Gore JM, Brophy CJ, Greenstone MA. How well do we care for patients with end stage chronic obstructive pulmonary disease (COPD)? A comparison of palliative care and quality of life in COPD and lung cancer. Thorax. Dec 2000;55(12):1000-1006.


Hall P, Schroder C, Weaver L. The last 48 hours of life in long-term care: a focused chart audit. Journal of the American Geriatrics Society. Mar 2002;50(3):501-506.


Hickman SE, Tilden VP, Tolle SW. Family reports of dying patients' distress: The adaption of a research tool to assess global symptom distress in the last week of life. Journal of pain and symptom management. 2001;22(1):565-574.


Levenson JW, McCarthy EP, Lynn J, Davis RB, Phillips RS. The last six months of life for patients with congestive heart failure. Journal of the American Geriatrics Society. 2000;48(5 Suppl):S101-S109.


Loos C, Briancon S, Frimat L, Hanesse B, Kessler M. Effect of end-stage renal disease on the quality of life of older patients. Journal of the American Geriatrics Society. Feb 2003;51(2):229233.


Lynn J, Teno JM, Phillips RS, et al. Perceptions by family members of the dying experience of older and seriously ill patients. SUPPORT Investigators. Study to Understand Prognoses and

Running Title: Symptoms in the Last Weeks


Preferences for Outcomes and Risks of Treatments. Annals of internal medicine. Jan 15 1997;126(2):97-106. 11.

McCarthy EP, Phillips RS, Zhong Z, Drews RE, Lynn J. Dying with cancer: Patients' function, symptoms, and care preferences as death approaches. Journal of the American Geriatrics Society. 2000;48(5 Suppl):S110-S121.


Mercadante S, Casuccio A, Fulfaro F. The course of symptom frequency and intensity in advanced cancer patients followed at home. Journal of pain and symptom management. Aug 2000;20(2):104-112.


Morita T, Tei Y, Inoue S. Impaired communication capacity and agitated delirium in the final week of terminally ill cancer patients: prevalence and identification of research focus. Journal of pain and symptom management. Sep 2003;26(3):827-834.


Nordgren L, Sorensen S. Symptoms experienced in the last six months of life in patients with end-stage heart failure. European journal of cardiovascular nursing : journal of the Working Group on Cardiovascular Nursing of the European Society of Cardiology. Sep 2003;2(3):213217.


Power D, Kearney M. Management of the final 24 hours. Irish medical journal. Sep 1992;85(3):93-95.


Ross DD, Alexander CS. Management of common symptoms in terminally ill patients: Part II. Constipation, delirium and dyspnea. American family physician. Sep 15 2001;64(6):1019-1026.


Semionov V, Singer Y, Shvartzman P. Prevalence and management of symptoms during the last month of life. The Israel Medical Association journal : IMAJ. Feb 2012;14(2):96-99.


Skilbeck J, Mott L, Page H, Smith D, Hjelmeland-Ahmedzai S, Clark D. Palliative care in chronic obstructive airways disease: a needs assessment. Palliative medicine. Jul 1998;12(4):245254.


The SUPPORT Principal Investigators. A controlled trial to improve care for seriously ill hospitalized patients. The study to understand prognoses and preferences for outcomes and risks

Running Title: Symptoms in the Last Weeks


of treatments (SUPPORT). The SUPPORT Principal Investigators. JAMA : the journal of the American Medical Association. Nov 22-29 1995;274(20):1591-1598. 20.

Tranmer JE, Heyland D, Dudgeon D, Groll D, Squires-Graham M, Coulson K. Measuring the symptom experience of seriously ill cancer and noncancer hospitalized patients near the end of life with the memorial symptom assessment scale. Journal of pain and symptom management. May 2003;25(5):420-429.


Teno JM, Clarridge BR, Casey V, et al. Family perspectives on end-of-life care at the last place of care.[see comment]. Journal of the American Medical Association. 2004;291(1):88-93.


Plonk WM, Jr., Arnold RM. Terminal care: the last weeks of life. Journal of palliative medicine. Oct 2005;8(5):1042-1054.


Pantilat SZ. End-of-life care for the hospitalized patient. The Medical clinics of North America. Jul 2002;86(4):749-770, viii.


Gilbertson-White S, Aouizerat BE, Jahan T, Miaskowski C. A review of the literature on multiple symptoms, their predictors, and associated outcomes in patients with advanced cancer. Palliative & supportive care. Mar 2011;9(1):81-102.


Solano JP, Gomes B, Higginson IJ. A comparison of symptom prevalence in far advanced cancer, AIDS, heart disease, chronic obstructive pulmonary disease and renal disease. Journal of pain and symptom management. Jan 2006;31(1):58-69.


Teunissen SC, Wesker W, Kruitwagen C, de Haes HC, Voest EE, de Graeff A. Symptom prevalence in patients with incurable cancer: a systematic review. Journal of pain and symptom management. Jul 2007;34(1):94-104.


Addington-Hall J, McPherson C. After-death interviews with surrogates/bereaved family members: some issues of validity. Journal of Pain & Symptom Management. 2001;22(3):784790.


Hinton J. How reliable are relatives' retrospective reports of terminal illness? Patients and relatives' accounts compared. Soc Sci Med. Oct 1996;43(8):1229-1236.

Running Title: Symptoms in the Last Weeks



Klinkenburg M, Willems DL, van der Wal G, Deeg DJH. Symptom Burden in the Last Week of Life. Journal of Pain Symptom Management. 2004;27(1):5-13.


Greenhalgh T, Peacock R. Effectiveness and efficiency of search methods in systematic reviews of complex evidence: audit of primary sources. BMJ. 2005-11-03 00:00:00 2005;331(7524):10641065.


von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. Oct 20 2007;370(9596):1453-1457.


Cooley ME. Symptoms in Adults with Lung Cancer: A Systematic Research Review. Journal of pain and symptom management. 2000;19(2):137-153.


Bailey FA, Burgio KL, Woodby LL, et al. Improving processes of hospital care during the last hours of life. Archives of internal medicine. Aug 8-22 2005;165(15):1722-1727.


Cohen LM, Germain M, Poppel DM, Woods A, Kjellstrand CM. Dialysis discontinuation and palliative care. American journal of kidney diseases : the official journal of the National Kidney Foundation. Jul 2000;36(1):140-144.


Goodlin SJ, Winzelberg GS, Teno JM, Whedon M, Lynn J. Death in the hospital. Archives of internal medicine. Jul 27 1998;158(14):1570-1572.


Zambroski CH, Moser DK, Roser LP, Heo S, Chung ML. Patients with heart failure who die in hospice. American heart journal. Mar 2005;149(3):558-564.


Conill C, Verger E, Henriquez I, et al. Symptom prevalence in the last week of life. Journal of pain and symptom management. Dec 1997;14(6):328-331.


Fainsinger R, Miller MJ, Bruera E, Hanson J, Maceachern T. Symptom control during the last week of life on a palliative care unit. Journal of palliative care. Spring 1991;7(1):5-11.


Hermann C, Looney, S. The effectiveness of symptom management in hospice patients during the last seven days of life. Journal of Hospice and Palliative Nursing. July-September 2001 2001;3(3):88-97.

Running Title: Symptoms in the Last Weeks



Lau KS, Tse DMW, Chen TWT, Lam PT, Lam WM, Chan KS. Comparing noncancer and cancer deaths in Hong Kong: A retrospective review. Journal of pain and symptom management. 2010;40(5):704-714.


Morita T, Ichiki T, Tsunoda J, Inoue S, Chihara S. A prospective study on the dying process in terminally ill cancer patients. The American journal of hospice & palliative care. Jul-Aug 1998;15(4):217-222.


Morita T, Tsunoda J, Inoue S, Chihara S. Risk factors for death rattle in terminally ill cancer patients: a prospective exploratory study. Palliative medicine. Jan 2000;14(1):19-23.


Paice JA, Muir JC, Shott S. Palliative care at the end of life: Comparing quality in diverse settings. American Journal of Hospice & Palliative Care. 2004;21(1):19-27.


Lichter I, Hunt E. The last 48 hours of life. Journal of palliative care. Winter 1990;6(4):7-15.


Casarett DJ, Inouye SK. Diagnosis and management of delirium near the end of life. Annals of internal medicine. Jul 3 2001;135(1):32-40.


Begg C, Cho M, Eastwood S, et al. Improving the quality of reporting of randomized controlled trials. The CONSORT statement. JAMA : the journal of the American Medical Association. Aug 28 1996;276(8):637-639.


Block SD. Psychological issues in end-of-life care. Journal of palliative medicine. Jun 2006;9(3):751-772.


Edwards A, Pang N, Shiu V, Chan C. The understanding of spirituality and the potential role of spiritual care in end-of-life and palliative care: a meta-study of qualitative research. Palliative medicine. Dec 2010;24(8):753-770.


Selman L, Beynon T, Higginson IJ, Harding R. Psychological, social and spiritual distress at the end of life in heart failure patients. Current opinion in supportive and palliative care. Dec 2007;1(4):260-266.


Williams AL. Perspectives on spirituality at the end of life: a meta-summary. Palliative & supportive care. Dec 2006;4(4):407-417.

Reviewer 2

Initial Search

Titles & Abstracts Reviewed n=562

Titles & Abstracts Reviewed n=446

Abstract Review

Abstracts Reviewed In-depth n=252

Abstracts Reviewed In-depth n=243

Article Screening

Articles Ordered & Screened n=89

Articles Ordered & Screened n=77

Joint Evaluation

Reviewer 1

Apply Inclusion & Exclusion Criteria n=25

Final Sample


Detailed Article Evaluation




Figure 1: Phases of Systematic Review

Quality Review n=13

Disagreements resolved by consensus

Additional Exclusions n=1 Sample Analyzed n=12

Table 1 – Databases and Search Terms Databases

Search terms

PubMed, Cumulative Index of

(death OR “terminal care” OR “palliative care” OR “end of life” OR

Nursing and Allied Health

dying OR “dying process” OR hospice OR “hospice care”)

Literature (CINAHL),


PsychINFO, and Web of

(signs OR symptoms OR “signs and symptoms” OR “signs and symptoms


respiratory” OR “signs and symptoms digestive” OR “neurobehavioral manifestations” OR “neurologic manifestations” OR “skin manifestations” OR “behavioral symptoms” OR “affective symptoms” OR depression)

Table 2: Summary of Studies Examining Signs and Symptoms in Last Two Weeks of Life Studies that restricted data collection of signs and symptoms Design, Data Source

Setting and

& Symptom Collection



Conill, C.,

• Prospective,


Diagnoses Included


Symptoms and Reported Prevalence

• Cancers

Last 7 days

18 Total

Verger, E.,

multicenter descriptive

home and

• Anorexia 80.1%

Henriquez, I.,



• Anxiety 45.5%

Saiz, N., Espier, • Patient interview

hospice unit

• Asthenia 81.8%

M., & Lugo, F.


• Confusional status 68.2%



• Constipation 55.1% • Cough 17.6% • Depression 38.6% • Diarrhea 9.1% • Dysphagia 46.0% • Dyspnea 46.6% • Dysuria 6.8% • Dry mouth 69.9%

• Hemorrhage 11.9% • Nausea 13.1% • Pain 30.1% • Paralysis 32.4% • Sleep disturbances 28.4% • Vomiting 10.2% Goodlin, SJ, Winzelberg, SG, Teno, JM,

• Retrospective descriptive study • Chart abstraction


• Cancer


• Ischemic heart


disease • Chronic

Last 2 days

10 Total • Agitation/Delirium/Restlessness 51% • Anorexia 19% • Anxiety 26%

Whedon, M, &

VA hospital

Lynn, J. (1998).



• Bed bound 68%



• Confusion 34%

• Infectious processes

• Dyspnea 51% • Fever 28%

• Stroke

• Incontinence 18%

• Congestive heart

• Nausea/Vomiting 19%


• Pain 59%

• Liver disease • Renal failure • AIDS • Dissecting aortic aneurysm • Valvular heart disease • Trauma • Other Hall, P, Schroder, C, & Weaver, L. (2002).

• Retrospective descriptive study • Chart abstraction

Long Term Care facilities

• Cognitive impairment

Last 48

7 Total


• Delirium 29%


• Cardiac disease

• Dysphagia 28%


• Respiratory

• Dyspnea 62%

disease • Neurologic disorders • Musculoskeletal

• Fever 24% • Myoclonus 16% • Pain 44% • Respiratory secretions 39%

• Cancer • Other Morita, T,

• Prospective


• Cancers


5 Total

Ichiki, T,

descriptive cohort


death” -

• Cyanosis of the extremities 80%

Tsunoda, J,


Care Unit

Last 24

• Mandibular respirations 95%



• Respiratory Secretions 35%

Inoue, S, &

• Chart abstraction

Chihara, S.

• Time to death after pulselessness at radial


(1998). Morita, T,

artery 2.6 +/- 4.2 hours • Prospective

Tsunoda, J,

descriptive study

Inoue, S, &

• Chart abstraction

Chihara, S.


• Cancers

Hospice Unit

Last 24

1 Total


• Respiratory Secretions 72.4%

Japan n=245

(2000). Paice, JA, Muir, C., & Shott, C. (2004).

• Retrospective descriptive study • Chart abstraction


• Cancers

Last 48

10 Total


• Cardiac &


• Agitation 32%



• Confusion 86%



• Constipation 32%

n = 195

• Bleeding

• Decubitus ulcers 26%

• Cirrhosis

• Dyspnea 78%


• Fatigue 46%

• Neuromuscular

• Fever 78%

• Other

• Pain 77% • Pulmonary congestion 92% • Sedation 36%

Studies that did not restrict data collection of signs and symptoms Design, Data Source

Setting and

& Symptom Collection



Diagnoses Included


VA Medical

• Cancer

Last 7 days

• Neurologic

Bailey, AF,

• Single group, before-

Burgio, KL,

after interventional


Woodby, LL,



13 Total - Pre/Post • Agitation/Delirium/Restlessness 23.1%/37.9%


• Heart disease

• Anorexia: 10.2%/44.2%

Redden, DT, et

Pre: n=108

• Lung disease

• Anxiety 2.8%/13.7%

al. (2005).

Post: n=95

• Liver disease

• Constipation 2.8%/8.4%

• Other

• Cough 10.2%/49.5%

Williams, BR,

• Chart abstraction


Symptoms and Reported Prevalence

• Depression 0.9%/3.2% • Dyspnea 30.6%/77.9% • Incontinence 16.7%/52.6% • Nausea/Vomiting 8.3%/18.9% • Pain 28.7%/57.9% • Skin problems 15.7%/31.6% • Sleep problems 4.6%/3.2% • Weakness 13.9%/49.5% Cohen, LM,

• Prospective


Germain, M,

observational cohort

clinics in

Popper, DM,


USA (6)

Woods, A, & Kjellstrand, CM. (2000).

• Chart abstraction

• End stage renal disease

Last 24 hours

8 Total – PresentAgitation/Delirium/Restlessness 30.4%

Canada (2)

• Diarrhea 13.9%


• Dysphagia 13.9% • Dyspnea 25.3% • Fever 20.3% • Myoclonus 27.8% • Nausea/Vomiting 12.7%

• Pain 41.7% Fainsinger, R,

• Retrospective


• Cancers

Last 7 days

4 Total

descriptive study


• Agitation/Delirium/Restlessness 39.0%

Bruera, E.

• Chart abstraction

Care Unit

• Dyspnea 46.0%


• Unrestricted


• Nausea/Vomiting 71%


• Pain 99.0%

Miller, MJ, &

n=100 Hermann, C & • Retrospective

Home hospice • Cancers

Looney, S.



descriptive study • Chart abstraction • Unrestricted

n = 100

• Cardiovascular disease • Chronic

Last 7 days

27 Total • Anxiety 18% • Bladder incontinence 36% • Bowel incontinence 31%


• Candidiasis 4%


• Confusion 46%


• Constipation 39%

• Other

• Decubitus 32%

• Unknown

• Dehydration 11% • Diarrhea 8%

• Dysphagia 50% • Dyspnea 70% • Edema 25% • Falls 3% • Fever 16% • Hallucinations 1% • Increased sleeping 42% • Insomnia 14% • Itching 3% • Jerking/Twitching 6% • Lethargy 68% • Mottling 19% • Mouth sores 2% • Nausea/Vomiting 22% • Pain 90% • Seizures 6% • Terminal congestion 47%

• Terminal restlessness 34% Lau, KS, Tse,

• Retrospective


Last 2

20 Total: (CRF/COPD/CHF/Cancer)

failure (CRF)


• Abdominal swelling/Ascites

DM, Chen, TW,

descriptive study

Lam, PT, Lam,

• Chart abstraction

Hong Kong

WM, & Chan,

• Unrestricted






disease (COPD)

KS. (2010).


• Chronic renal

Non-cancer: n=656: CRF=239

• Chronic

• Congestive heart failure (CHF) • Cancer

7.9%/2.9%/2.3%/26.8% • Agitation/Delirium/Restlessness 6.7%/1.2%/5.1%/13.0% • Anorexia 29.3%/16.1%/13.7%/63.8% • Anxiety 0.0%/1.2%/0.6%/5.8% • Cachexia 1.7%/5.4%/1.7%/27.5% • Confusion 14.6%/18.6%/9.7%/40.6%


• Cough 18.0%/49.2%/18.3%/31.2%


• Depression 1.7%/2.5%/2.3%/10.9% • Diarrhea 21.3%/14.5%/12.0%/42.8% • Dyspnea 53.1%/86.0%/65.1%/77.5% • Edema 38.5%/22.7%/49.7%/42.8% • Fatigue 28.5%/26.0%/18.9%/84.8% • Hemoptysis 0.8%/4.1%/1/1%/5.8%

• Nausea/Vomiting 18.8%/5.4%/6.3%/37.0% • Oral problems 2.5%/8.3%/2.9%/12.6% • Pain 34.3%/27.3%/28.6%/85.5% • Pruritus 2.1%/0.4%/0.0%/1.4% • Skin problems 10.4%/4.1%/13.1%/17.4% • Sleep problems 3.3%/21.1%/1.7%/21.7% • Urinary problems 10.5%/7.9%/8.6%/22.5% Zambroski,

• Retrospective

Home, Long

• Heart failure

Last 7 days

4 Total

descriptive study

Term Care

• Confusion 47.8%

DK, Roser, LP,

• Chart abstraction

facility, or

• Dyspnea 63.3%

Heo, S, &

• Unrestricted


• Edema 45.6%

Chung, ML.

hospice unit

• Incontinence 38.9%



CH, Moser,


Table 3: Weighted Prevalence of Signs and Symptoms by Study Restricted Conill Goodlin 1997


Hall 2002*


Morita Morita 1998


Paice 2004*

Bailey Bailey Cohen Fainsinger Hermann 2005


Signs &

# of














Abdominal swelling











Bed bound
















5.8 26.3




Bowel problems

















4.0 34.0

86.0 32.0

46.0 2.8





39.0 28.3


Cyanosis of extremities











Dry mouth










11.0 0.9


28.0 51.0





3.5 14.0



























42.0 28.0







16.0 1.0 2.6

11.9 18.0




Increased sleeping







Mouth sores




2.0 18.0



Nausea/ Vomiting










Oral problems






















Respirations with mandibular movement



Respiratory secretions











36.0 6.0

Skin integrity problems








Sleep problems














* = calculated from percentage

10.7 14.0


# pts with s/sx across studies

Weighted prevalence of s/sx (%) †

Unweighted prevalence of s/sx (%)

95% confidence interval of unweighted prevalence

Abdominal swelling Agitation Anorexia Anxiety Bed bound Bowel problems (not constipation or diarrhea) Cachexia Candidiasis

1 8*** 4*** 5*** 1 1 1 1

839 1805 1322 1422 104 839 839 100

67 376 436 153 71 166 58 4

7.99 20.83 32.98 10.76 68.27 19.79 6.91 4.00

7.99 31.39 36.16 17.90 68.00 19.79 6.90 4.00

21.89, 40.89 2.09, 70.22 0.69, 35.11 -

Confusion Constipation Cough Cyanosis of extremities

6 4*** 3*** 1

1504 674 1218 100

565 209 327 80

37.57 31.01 26.85 80.00

50.07 27.46 26.64 80.00

24.74, 75.40 0.48, 54.44 -0.07, 53.34 -

Dehydration Depression Diarrhea Dry mouth

1 3*** 3 1

100 1218 355 176

11 101 31 123

11.00 11.00 8.29 11.55 8.73 9.60 69.89 69.90

-17.20, 40.30 0.02, 19.18 -

Dysphagia Dyspnea Edema Falls

4 10*** 3 1

540 2071 1029 100

194 1285 357 3

35.93 34.50 62.05 56.10 34.69 34.30 3.00 3.00

7.95, 61.04 44.06, 68.14 11.90, 56.69 -





38.01 47.83

3.18, 92.48

Fever Hallucinations Hemoptysis

5 1 1

663 100 839

257 1 22

38.76 33.20 1.00 1.00 2.62 2.62

1.61, 64.78 -





11.93 11.90


# studies assessing s/sx*

# pts assessed across studies**

Table 4: Comparison of Weighted and Unweighted Prevalence of Signs and Symptoms

Signs & Symptoms (s/sx)

Incontinence Increased sleeping Mouth sores

4*** 1 1

497 100 100

187 42 2

37.63 38.28 42.00 42.00 2.00 2.00

11.18, 65.37 -





16.76 17.33






19.43 23.74

7.25, 40.22

Oral problems Pain Paralysis Pruritus

1 10*** 1 2

839 2071 176 939

54 978 57 11

6.44 6.44 47.22 52.42 32.39 32.40 1.17 2.00

35.22, 69.62 -10.72, 35.33

Respirations with mandibular movement




95.00 95.00


Respiratory secretions




53.33 51.40

22.64, 80.15

Sedation Seizures Skin integrity problems Sleep problems Urinary problems Weakness * Total number of studies = 12

1 1 4*** 4*** 2 2***

195 100 1337 1318 1015 379

70 6 211 118 102 206

35.90 6.00 15.78 8.95 10.05 54.35

35.90 6.00 23.03 10.72, 35.33 11.16 -1.90, 24.21 8.76 -16.20, 33.73 23.89 -31.66, 79.45

** Total number patients represented across studies = 2416 *** Includes Bailey, which has pre- and post- intervention results calculated separately † = [(# pts with s/sx across studies)/(# pts assessed across studies)]*100 ‡ = (% with s/sx across studies)/(# studies assessing s/sx)

Suggest Documents