If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Prior work has shown that symptoms leading to restrictions in daily activities are common at the end of life. Hospice is a Medicare benefit designed to alleviate distressing symptoms in the last 6 months of life. The effect of hospice on the burden of such symptoms is uncertain.
From an ongoing cohort study of 754 community-dwelling older persons, aged ≥70 years, we evaluated 241 participants who were admitted to hospice from March 1998 to December 2013. A set of 15 physical and psychological symptoms leading to restricted activity (ie, cut down on usual activities or spend at least half the day in bed) were ascertained during monthly telephone interviews in the year before and 3 months after hospice admission.
The prevalence and mean number of restricting symptoms increased progressively until about 2 months before hospice admission, before increasing precipitously to a peak around the time of hospice admission. After the start of hospice, both the prevalence and the mean number of restricting symptoms dropped markedly. For several symptoms deemed most amenable to hospice treatment, including depression and anxiety, the prevalence dropped to levels comparable to or lower than those observed 12 months before the start of hospice. The trends observed in symptom prevalence and mean number of symptoms before and after hospice did not differ appreciably according to hospice admission diagnosis or sex. The median duration of hospice (before death) was only 15 days.
The burden of restricting symptoms increases progressively several months before the start of hospice, peaks around the time of hospice admission, and decreases substantially after the start of hospice. These results, coupled with the short duration of hospice, suggest that earlier referral to hospice may help to alleviate the burden of distressing symptoms at the end of life.
One of the most common approaches to ameliorating distressing symptoms is palliative care. Prior studies have shown that palliative care can significantly improve quality of life and even increase survival rates.
At the end of life, hospice, a specific type of palliative care, is often considered. Hospice use has grown substantially in the past decade. However, the median length of stay in hospice has remained very short, at approximately 2.5 weeks.
this short duration raises concerns about whether hospice is being properly utilized.
Although the literature is replete with studies on hospice, relatively little is known about the course of distressing symptoms before and after the start of hospice. Prior studies of distressing symptoms at the end of life have focused on specific subgroups, such as patients with cancer,
we used a longitudinal cohort design to study symptom burden in the last year of life. To our knowledge, comparable longitudinal data do not exist for patients before and after admission to hospice.
In the present study, we set out to evaluate the course of symptoms before and after hospice admission. We were particularly interested in determining whether the occurrence of restricting symptoms (ie, those that lead to bed rest or cause one to cut down on activities) is reduced after the start of hospice. To address this question, we used data from a unique longitudinal study that includes monthly assessments of restricting symptoms before and after the start of hospice. Given the wide range of symptom severity, we focused on restricting symptoms because they are the most likely to be important for patients and their caregivers. We categorized these symptoms on the basis of clinical judgment into three groups, from most to least amenable to hospice treatment. Given previously reported differences in hospice referral, utilization, and functional decline at the end of life,
involving 754 older persons. Eligible participants were community-dwelling, nondisabled, and aged ≥70 years at the time of enrollment. Exclusion criteria were substantial cognitive impairment with no available proxy, life expectancy <12 months (at the time of eligibility assessment), plans to move out of the area, or inability to speak English. Of the 1002 persons eligible for the study, 754 agreed to participate and were enrolled between March 1998 and October 1999. The research protocol was approved by the Yale Human Investigation Committee.
We identified 260 hospice admissions through 2013 using Medicare claims data and review of medical records.
Of these, 12 had previously dropped out of the study and 1 had died before their first monthly interview. To permit pre–post comparisons, 6 additional participants were excluded because they did not have at least 6 months of symptom data before hospice admission, leaving 241 participants in the analytic sample.
Comprehensive home-based assessments were completed at baseline and subsequently at 18-month intervals, whereas telephone interviews were completed monthly. During the comprehensive assessments, data were collected on demographic characteristics, depressive symptoms (Center for Epidemiological Studies Depression Scale
examination score <24), and the presence of 9 self-reported, physician-diagnosed comorbid conditions: hypertension, myocardial infarction, heart failure, stroke, cancer, diabetes mellitus, hip fracture, arthritis, and chronic lung disease. Body mass index was calculated using participants' self-reported height and weight. For descriptive purposes and to fulfill federal regulations regarding the inclusion of minority participants in studies funded by the US National Institutes of Health, participants were asked to identify their race and ethnicity.
Hospice admission diagnosis codes were obtained from the Medicare claims data.
These diagnoses were classified into 7 categories (Table 1) based on the Centers for Medicare & Medicaid Services International Classification of Diseases, 9th Revision diagnosis codes. For 3 participants without Centers for Medicare & Medicaid Services Medicare claims data, the hospice admission diagnoses were inferred from information obtained from death certificates, which were coded by a certified nosologist.
Table 1Hospice Admission Diagnoses (N = 241)
Uncertain or unspecified neoplasm
Parkinson's and other cerebral degenerations
Aneurysm or aortic dissection
Respiratory abnormality not elsewhere classified
Gastrointestinal and liver disease
Adult failure to thrive
Debility not otherwise specified
ICD-9 = International Classification of Diseases, 9th Revision.
First, participants were asked 2 questions related to restricted activity: “Since we last talked (ie, during the last month), have you stayed in bed for at least half a day due to an illness, injury, or other problem?” and “Since we last talked, have you cut down on your usual activities due to an illness, injury, or other problem?” Second, if participants answered “yes” to either of these questions, they were asked whether they had any of 24 prespecified symptoms and problems since the last interview.
Third, immediately after each “yes” response to a specific symptom or problem, participants were asked, “Did this cause you to stay in bed for at least half a day or to cut down on your usual activities?” During pilot testing among 20 persons, we found that the test–retest reliability (mean time between assessments, 4.1 days) of this protocol was high, with κ of 0.90 for the presence or absence of restricted activity and κ of ≥0.75 for the presence or absence of all restricting symptoms.
Given the aims of the present study, we chose to omit the restricting problems (eg, change in medications, death of a family member) and focus on the 15 restricting symptoms: fatigue; dizziness or unsteadiness; difficulty with memory or thinking; swelling in feet or ankles; cold or flu symptoms; musculoskeletal pain; dyspnea; depression; anxiety; poor or decreased vision; arm or leg weakness; nausea, vomiting, or diarrhea (determined through a single question); urinary symptoms; difficulty sleeping; and chest pain or tightness. These symptoms were categorized a priori on the basis of clinical judgment into 3 groups according to how amenable they might be to hospice treatment: most, intermediate, and least.
When participants were too ill or otherwise unable to complete the monthly interviews, proxy data were obtained using a standard protocol.
We required that proxies were cognitively intact and lived with the participant or saw the participant regularly. Of the 2629 monthly interviews in the present study, 41.8% were completed by a proxy respondent. To determine the accuracy of these reports, we had previously conducted interviews with 20 participants and their designated proxies (separately) for 6 months and compared their responses to questions assessing the occurrence of restricted activity. Concordance was substantial, with κ = 0.66 (95% confidence interval, 0.50-0.83).
Demographic and clinical characteristics of the analytic sample were summarized using information from the comprehensive assessment that immediately preceded the study period (ie, year before hospice). The prevalence of restricting symptoms was calculated by dividing the number of participants with any restricting symptom in a particular month by the total number of participants who completed an interview that month. Similarly, the monthly occurrence of specific restricting symptoms was calculated by dividing the number of participants with that symptom in a specific month by the total number of participants completing the interview. Finally, the mean number of restricting symptoms among participants was calculated. The monthly occurrence of restricting symptoms and the mean number of restricting symptoms were plotted in the 12 months before hospice admission, as well as 3 months after hospice admission. Follow-up beyond 3 months was not included because of the small number of non-decedents. Because restricting symptoms were ascertained “during the past month” and the mean number of days between hospice admission and the first postadmission interview was 14, many of the responses during the first postadmission interview likely refer to symptoms before the start of hospice. To evaluate this possibility, we stratified the prevalence and the number of restricting symptoms into 3 groups according to when the first postadmission interview was completed relative to the start of hospice (ie, 1-10, 11-20, >20 days).
To account for the high short-term mortality rate after admission to hospice, we also plotted the monthly occurrence of restricting symptoms before and after the start of hospice exclusively for participants who completed at least 1 post-hospice monthly interview. To address potential selection effects, we repeated these analyses among participants who completed at least 3 post-hospice monthly interviews. A final set of analyses focused on the monthly occurrence of restricting symptoms according to hospice admission diagnosis and sex.
All analyses were performed using SAS, version 9.4 (SAS Institute, Cary, NC).
As shown in Table 1, the most common admission diagnosis for hospice was cancer. Table 2 provides the characteristics of the analytic sample. The mean age was 88.4 years, with approximately 30% of the sample being younger than 85 years. Approximately two-thirds were female, and more than one-third had less than a high school education. Approximately 9 out of 10 were non-Hispanic white. Participants had a diverse set of chronic conditions, with an average of 2.6. The median duration of hospice was 15 days (interquartile range, 5-51 days). Only a minority of participants were admitted to an inpatient hospice facility (n = 67 or 27.8%). The others received hospice at home in the community (n = 93 or 38.6%) or a nursing home (n = 81 or 33.6%).
As shown in Figure 1, the prevalence of any restricting symptoms increased slowly until approximately 2 months before hospice admission, when it increased precipitously, reaching a peak shortly after the start of hospice. Subsequently the prevalence of any restricting symptoms dropped markedly from 0.69 to 0.48 before flattening out at 0.46. The pattern for the mean number of restricting symptoms was similar, with a peak of 3.5 restricting symptoms before dropping to 2.1 and flattening out at 2.0. Both the prevalence and number of restricting symptoms decreased as the number of days increased between the time of hospice admission and the time of the first post-hospice interview (Table 3).
Table 3Restricting Symptoms According to Timing of First Post-Hospice Interview (n = 111)
Days After Hospice Admission
Prevalence of Restricting Symptoms
Mean No. of Restricting Symptoms
A total of 130 participants died before their first post-hospice interview.
Figure 2 provides the results for each of the restricting symptoms, categorized into 3 groups, from most to least amenable to hospice treatment. For Figure 2A and B, the prevalence of most restricting symptoms followed the same general pattern as for any restricting symptoms, with a slow increase followed by a more pronounced increase shortly before the start of hospice and a subsequent decrease after the start of hospice. These results were most striking for fatigue, depression, anxiety (Figure 2A), and arm/leg weakness (Figure 2B). For several of these symptoms, the prevalence after admission to hospice dropped to levels comparable to those observed 12 months before the start of hospice. For some, including depression and anxiety, the prevalence 3 months after hospice admission was lower than 12 months before hospice admission (Figure 2A). In contrast, the prevalence of the 2 symptoms deemed least amenable to hospice treatment—cold/flu symptoms and poor eyesight—was low throughout the follow-up period, with little change before or after hospice admission (Figure 2C). These results were consistent among participants who had at least 1 post-hospice follow-up interview (Supplemental Figure, available online). Additional analyses among participants who had at least 3 post-hospice follow-up interviews showed similar patterns (results not shown).
Although there was some month-to-month variability, the trends observed for symptom prevalence and mean number of symptoms before and after hospice did not differ appreciably according to hospice admission diagnosis or sex (Figure 3). For both Figure 2 and Figure 3, the estimates at 3 months post-hospice are less stable than those at earlier time points because they are based on a relatively small proportion of participants surviving through 3 months. For example, the low values at 3 months for organ failure (Figure 3) are likely due to the small number of participants (n = 5) who survived to this point from the initial group of 36.
In this prospective cohort study of older persons, we found that the burden of restricting symptoms increases progressively several months before the start of hospice, peaks around the time of hospice admission, and decreases substantially after the start of hospice. These trends were observed most notably for fatigue, depression, anxiety, arm/leg weakness, and swelling in feet/ankles. As in prior studies,
median survival after hospice admission was only 15 days. When coupled with the short duration of hospice, the high burden of restricting symptoms before hospice admission and the reduction of these symptoms after the start of hospice suggest that earlier referral to hospice may help to alleviate the burden of distressing symptoms at the end of life.
Our results build on those of prior studies that have shown beneficial effects of palliative care on symptom burden.
In contrast, our study included older persons with a diverse array of hospice admission diagnoses and demonstrated that the observed pattern of restricting symptoms before and after the start of hospice did not differ according to these diagnoses. The availability of detailed information on symptom burden before hospice admission is a unique feature of our study.
we evaluated a wide range of symptoms, allowing us to identify a subset of symptoms, including fatigue, depression, anxiety, arm/leg weakness, and foot/ankle swelling, that seem to respond particularly well to hospice care. Our focus on symptoms leading to restricted activity enhances the clinical relevance of our findings because proper management of these symptoms may substantially improve quality of life while reducing caregiver burden. Although prior studies have demonstrated sex-specific differences in hospice referral and utilization,
we found no differences between men and women in the prevalence or mean number of restricting symptoms before and after the start of hospice.
The 1-month assessment intervals did not permit us to precisely identify when the burden of restricting symptoms peaked and subsequently declined. However, we found that both the prevalence and number of restricting symptoms decreased as the number of days increased between the time of hospice admission and the time of the first post-hospice interview, supporting the beneficial effects of hospice on the reduction of restricting symptoms, and suggesting that symptom burden likely peaked near the time of hospice admission.
Our results should be interpreted carefully because only 111 individuals, a large minority of participants, survived to complete a post-hospice interview. However, the results were unchanged when the analyses were limited to participants who completed the first post-hospice interview. Similarly, the patterns of restricting symptoms were comparable among the small proportion of participants who completed at least 3 post-hospice interviews, thereby diminishing the possibility of a selective survival effect. Although causality can be suggested but not proven by a cohort study, prior research has shown a marked and unperturbed increase in symptom burden in the last year of life in the absence of hospice.
palliative care should have reduced the prevalence and number of restricting symptoms before the start of hospice, suggesting that our results might be conservative. Because our study participants were members of a single health plan in a small urban area in the United States, our results may not be generalizable to older persons in other settings. However, the demographic characteristics of our cohort at the time of enrollment reflected those of older persons in New Haven County, Connecticut, which were similar to the characteristics of the US population as a whole, with the exception of race or ethnic group.
The generalizability of our results is further enhanced by our high participation, which was greater than 75%. Although a substantial minority of our interviews were completed by proxies, we have previously demonstrated a high concordance between our proxy and participant reports for restricted activity.
In summary, we found that the burden of restricting symptoms at the end of life decreases significantly after the start of hospice for older men and women, irrespective of hospice admission diagnosis. These results, coupled with the short duration of hospice, suggest that earlier referral to hospice may help to alleviate the burden of distressing symptoms at the end-of-life.
We thank Denise Shepard, BSN, MBA, Andrea Benjamin, BSN, Barbara Foster, and Amy Shelton, MPH, for assistance with data collection; Wanda Carr and Geraldine Hawthorne, BS, for assistance with data entry and management; Peter Charpentier, MPH, for design and development of the study database and participant tracking system; and Joanne McGloin, MDiv, MBA, for leadership and advice as the Project Director.
Funding: The work for this report was funded by a grant (R37AG17560) from the National Institute on Aging and by the Richard Alan Hirshfield Memorial Fellowship. The study was conducted at the Yale Claude D. Pepper Older Americans Independence Center (P30AG21342). TMG is the recipient of an Academic Leadership Award (K07AG043587) from the National Institute on Aging.
Conflict of Interest: None.
Authorship: All authors had access to the data and a role in the writing the manuscript.