Abstract
Context:
Given that the composition of hospice patients’ terminal diagnoses has become increasingly diverse, understanding whether hospices provide quality care to patients, regardless of disease, is important. However, data comparing diagnosis and caregiver-reported outcomes remain scarce.
Objectives:
To analyze the association between the composition of patients’ terminal diagnoses and caregiver-reported quality measures.
Methods:
Using cross-sectional, publicly available data from the Centers for Medicare & Medicare Services (CMS), we analyzed data collected from 2015–2019. We conducted general linear model analyses to identify associations between hospice characteristics/practices and caregiver-reported outcomes.
Results:
Of the 2810 hospices, those that cared for a greater percentage of dementia patients had fewer caregivers, on average, who rated hospice a 9 or 10 (where 0 = low, 10 = high; β = −.094; 95% CI = −.147, −.038), reported they always received help for pain and symptoms (β = −.106, CI = −.156, −.056), and reported definitely having received the training they needed (β = −.151, CI = −.207, −.095). Those caring for more stroke patients had fewer caregivers, on average, who rated hospice a 9 or 10 (β = −.184, CI = .252, −.115), reported they always received help for pain and symptoms (β = −.188, CI = −.251, −.126), reported definitely having received the training they needed (β = −.254, CI = −.324, −.184), and reported that the hospice offered the right amount of emotional/spiritual support (β = −.056, CI = −.093, −.019).
Conclusion:
Hospices that cared for a greater proportion of dementia and stroke patients had poorer scores on caregiver-reported quality measures. These findings support efforts to identify mechanisms underlying these differences and to design strategies to ensure optimal outcomes for hospice patients regardless of diagnosis.
Keywords: hospice, quality, caregiver, end-of-life care
Introduction
When the Medicare hospice benefit began in the United States in 1982, 95% of hospice patients had a terminal cancer diagnosis.1 Over the past 4 decades, hospices have expanded to provide end-of-life care to patients with a diverse range of terminal conditions. In 2018, terminal cancer (29.6%), heart disease (17.4%), and dementia (15.6%) were the top 3 admitting diagnoses of hospice patients.2,3
Given the diversity of terminal medical diagnoses in hospice care, it is important to understand whether and how the composition of patients served by hospice agencies is associated with quality indicators. Hospice care was originally developed with the cancer patient in mind. Understanding whether hospice care delivery and organizational practices are providing similar quality of care to all patients, regardless of diagnosis, is crucial for identifying and addressing potential inequalities and gaps in end-of-life care.
In particular, caring for a patient with a terminal diagnosis of dementia introduces unique challenges. Dementia patients admitted to hospices tend to have longer lengths of stay and make up a greater proportion of live discharges,4,5 reaching up to 39% of this population.6 Length of stay and live discharge are two measures that the Centers for Medicare & Medicare Services (CMS) track as potential indicators of concern, as they could influence hospice practices regarding resource allocation (e.g., services for patients and caregivers, staff training) and/or patient selection, potentially impacting care outcomes. In addition, live discharge is considered a burdensome transition for family members that may influence their perceptions of hospice care quality,7 which could in turn influence caregiver ratings of hospices. While a 2007 study by Mitchell et al. found that overall hospice quality of care was similar among hospice patients with dementia, cancer, and other chronic diseases, the authors noted general problems in care, including providing emotional support and communicating information about the patient’s condition.8
From a clinical standpoint, patients with a terminal diagnosis of dementia are often unable to verbalize their care needs and can exhibit distressing behavioral sequelae (e.g., agitation, delirium). In a qualitative study conducted by Harrison et al, the authors examined hospice staff perspectives when caring for patients with dementia and noted disease-specific care challenges (e.g., challenges in coordinating care decisions with caregivers).9 Given the factors that can influence how hospice care is delivered to dementia patients and potentially to patients with other terminal diagnoses, it is important to understand whether the quality of care that is delivered is impacted by the composition of patients’ terminal diagnosis.
The primary objective of this study is to analyze the association between the composition of patients’ terminal diagnosis (i.e., percentage of patients with cancer, heart disease, dementia, respiratory disease, stroke, or other conditions) in hospice organizations and caregiver-reported hospice ratings (i.e., the proportion of caregivers that rated the hospice highly, defined as a score of 9 or 10 where 0 = low and 10 = high), collected via the Consumer Assessment of Healthcare Providers and Systems (CAHPS) Hospice Survey. The secondary objective is to assess for associations between the composition of patients’ terminal diagnoses and 3 additional salient caregiver-reported measures (i.e., providing help for patient’s pain and symptoms, providing adequate training, and providing spiritual and emotional support), which are important tenets of providing quality end-of-life care.10,11
Methods
Study Design
This cross-sectional study used public data from CMS12 and was approved by the institutional review board of Weill Cornell Medicine.
Data Sources
Data were obtained from 3 CMS datasets for this analysis. (1) Hospice – General Information (HGI), provides information on hospice ownership type (data collected from 2015–2019); (2) Hospice care – Provider CAHPS Hospice Survey Data (HCPC), which reports the score of each hospice on the caregiver-reported quality measures (data collected from 2018–2019); (3) Hospice – Provider Data (HPD), includes information on hospice characteristics and practices (data collected from 2015–2019), as well as each hospices’ patient composition by diagnosis (data collected during 2017). The hospice practices include percentage of patients who received the comprehensive assessment at admission (reported by hospices) and percentage of patients visited within 72 hours of death (derived from Medicare claims data).
Study Sample Selection
The initial cohort from the 3 datasets consisted of 5041 hospices. We excluded all hospice organizations that were flagged by CMS with annotations that indicated: (1) hospice data were not reported because the organization was too new or too small to be required to participate in the survey, or no patients at the hospice met the criteria for the measures for the reporting period (n=2006); (2) hospice did not report data for the 2015–2019 period (n = 117); (3) hospice did not meet the required minimum set for public reporting (n = 103); and/or (4) hospice did not submit scores for the reporting period (n = 5). The final analytic sample included data from 2810 agencies.
Outcome Variables
We examined the 4 key outcomes of interest, all reported by family caregivers, analyzing the highest rated answer for each measure: (1) percentage of caregivers who rated the hospice a 9 or 10 (0 = low, 10=high), (2) percentage of caregivers who reported that the patient “always” received help for pain and symptoms, (3) percentage of caregivers who reported “definitely” having received the training they needed to care for patient, and (4) percentage of caregivers who felt they received the “right amount” of emotional and spiritual support from hospice staff. We selected these outcomes based on the Clinical Practice Guidelines for Quality Palliative Care that define the goals of hospice care as providing patients with optimal medical care and pain management, supporting the primary caregiver, and delivering emotional and spiritual support.10
Independent Variables
These variables included the composition of patients cared for by each hospice organization (i.e., percentage of patients with cancer, heart disease, dementia, respiratory disease, stroke, or other conditions). The data provides composition of patients at the hospice level and not at the individual level. Therefore, we were unable to analyze responses by grouping terminal diagnoses together and comparing them directly with one another.
We also examined 4 hospice characteristics: ownership type, U.S. region, average daily census, and year of hospice Medicare certification. We recategorized the ownership type variable that originally had 5 categories (profit, non-profit, government, combination government and non-profit, and other) to 3 (profit, non-profit, and other) for parsimony. Using the census for regions of the U.S.,13 we created a variable grouping hospices by their geographical region (i.e., South, Midwest, West, and Northeast).
Lastly, we included 2 hospice practice variables. The first was the percentage of patients who received a comprehensive assessment at admission.14 The comprehensive assessment at admission is based on 7 measures of the Hospice Item Set: (1) screened for pain, (2) assessed/managed the pain, (3) screened for dyspnea, (4) assessed/managed the dyspnea, (5) treated with opioids and assessed/managed the constipation, (6) beliefs and values addressed, and (7) life-sustaining treatment preferences discussed.14 The second was the percentage of patients in each hospice who received visits within 72 hours of death.
Statistical Analysis
We obtained descriptive statistics to examine the overall composition of hospice patients by terminal diagnosis along with other independent and outcome variables. The primary analyses were by general linear models that included a number of independent variables. Specifically, we included the composition of patients cared for by each hospice organization (i.e., percentage of patients with cancer, heart disease, dementia, respiratory disease, stroke, or other conditions) in this model a priori, given that our objective was to analyze the composition of hospices by diagnosis. We also included ownership type and U.S. region as classification factors; and average daily census and certification year (covariates) in the model a priori, given supporting literature showing their associations with hospice quality ratings.15 Lastly, we included the 2 hospice practice variables to the model (i.e., percentage of patients who received a comprehensive assessment at admission and percentage of patients in each hospice who received visits within 72 hours of death) a priori as they were seen as important variables that may impact the outcome measures.
Statistical analysis was conducted using STATA BE version 14.0 (College Station, Texas).
Results
Description of Study Cohort
Table 1 shows data on hospice characteristics, hospice practices, and patient composition. Most hospices were for-profit (n = 1,578, 56%), followed by non-profit (n = 828, 30%) and other (n = 404, 14%). The mean daily census was 94 patients. With regard to the hospice practice variables, the mean percentage of patients across hospices (n = 2758) who received a comprehensive assessment at admission was 91.7%. On average, while 87.4% of patients were visited by a hospice team member within 72 hours of their death, cancer (26%) made up the largest portion of patients in the sample, followed by dementia (20%) and heart disease (18%). Table 2 summarizes the caregiver-reported hospice quality measures. For the primary outcome, the average percent of caregivers who rated hospice a 9 or 10 (out of 10) was 81%, followed by 14% with a rating of 7 or 8, and 5% with a rating of 6 or lower.
Table 1.
Descriptive Dataa for Hospice Characteristics, Practices, and Patient Composition.
| Independent variable (N) | |
|---|---|
| Hospice characteristics | |
| Ownership type (2,810) | |
| Profit | 1578 (56%) |
| Non-profit | 828 (29%) |
| Otherb | 404 (14%) |
| Hospices by region (2,810) | |
| South | 1097 (39%) |
| Midwest | 761 (27%) |
| West | 593 (21%) |
| Northeast | 359 (13%) |
| Average daily census (2,786) | |
| Mean (SD) | 94 (184) |
| Median | 56 |
| Certification year (2,810) | |
| Mean (SD) | 2000 (10) |
| Median | 2002 |
| Hospice practices | |
| Received comprehensive assessment at admission % (2758) | |
| Mean (SD) | 91.7 (10.2) |
| Visited within 72 hours of death % (2752) | |
| Mean (SD) | 87.4 (11.8) |
| Primary diagnosis on admission | |
| Cancer % (2722) | |
| Mean (SD) | 26.0 (9.5) |
| Dementia % (2539) | |
| Mean (SD) | 20.1 (9.8) |
| Heart disease % (2501) | |
| Mean (SD) | 18.3 (6.9) |
| Respiratory disease % (2177) | |
| Mean (SD) | 9.9 (4.0) |
| Stroke % (2074) | |
| Mean (SD) | 9.7 (5.4) |
| Other terminal conditions % (1690) | |
| Mean (SD) | 7.1 (3.9) |
Data obtained from the Centers for Medicare & Medicare Services.
Government, combination government and non-profit, and other ownership types.
N = number of hospices; SD = standard deviation.
Table 2.
Scoresa of the Analyzed CAHPS Hospice Survey Caregiver-reported Quality Measures.
| Outcome (N) | Percent of Caregivers Who Provided That Outcome, Displayed as Mean for All Agencies (SD) |
|---|---|
| Primary outcome | |
| Caregiver rating of hospice (2,810) | |
| 9 or 10 | 81.0% (6.4) |
| 7 or 8 | 14.2% (4.6) |
| 6 or lower | 4.8% (3.0) |
| Secondary outcomes | |
| Patients got help for pain and symptoms (2,810) | |
| Always | 75.1% (6.1) |
| Usually | 15.2% (4.3) |
| Sometimes/never | 9.7% (3.2) |
| Caregiver received needed training by hospice (2,797) | |
| Definitely | 75.5% (7.0) |
| Somewhat | 15.2% (4.4) |
| Did not | 9.3% (3.9) |
| Amount of emotional and spiritual support hospice provided (2,810) | |
| Right amount | 90.2% (3.5) |
| Not the right amount | 9.8% (3.5) |
Data obtained from the Centers for Medicare & Medicare Services.
CAHPS = Consumer Assessment of Healthcare Providers and Systems; N= number of hospices; SD = standard deviation.
Percentage of Caregivers who Rated Hospice a 9 or 10
Table 3 shows the multivariable analyses for the primary outcomes. Hospices with a higher proportion of patients with a dementia (P < .01) or stroke (P < .001) diagnosis were associated with a lower percentage of caregivers who rated a hospice a 9 or 10; caregiver rating was not associated with the proportion of patients with a diagnosis of cancer, heart disease, respiratory disease, or other conditions.
Table 3.
Estimated Regression Coefficients for the Multivariable Analysis of Independent Variables With Achieving the Highest/Best Score on 4 Caregiver-reported Quality Outcomes in the CAHPS Hospice Survey.
| Independent Variable | Outcome (Based on % of Caregivers Who Reported the Measure) | |||
|---|---|---|---|---|
| Caregiver Rating Hospice a 9 or 10 (95% CI) | Patients Always Received Help for Pain and Symptoms (95% CI) | Caregivers Definitely Received the Training They Needed (95% CI) | Hospice Offered the Right Amount of Emotional and Spiritual Support (95% CI) | |
| Hospice patient composition by primary diagnosis | ||||
| Cancer | −.051 (−.114, .013) | −.027 (−.085, .0312) | .038 (−.027, .103) | −.074d (−.108, −.039) |
| Dementia | —.094c (−.147, −.038) | −.106d (−.156, −.056) | −. 15ld (−.207, −.095) | −.026 (−.056, .003) |
| Heart disease | −.019 (−.076, .038) | −.024 (−.077, .028) | −.043 (−.102, .015) | −.027 (−.058, .004) |
| Respiratory disease | −.024 (−.111, .063) | .028 (−.052, .108) | .017 (−.072, .106) | −.028 (−.075, .019) |
| Stroke | —.184d (.252, −.115) | −.188d (−.251, −.126) | −.254d (−.324, −.184) | −.056d (−.093, −.019) |
| Other conditions | −.084 (−.169, .000) | −.1344c (−.212, −.056) | −.114b (−.200, −.027) | −.074d (−.120, −.028) |
| Hospice characteristics | ||||
| Ownership type | ||||
| For-profit | Ref | Ref | Ref | Ref |
| Non-profit | l.988d (1.270, 2.706) | .806b (.145, 1.466) | .229 (−.506, .964) | .980d (.591, 1.370) |
| Otherb | .896b (.0563, 1.736) | .613 (−.160, 1.386) | .447 (−.413, 1.306) | .358 (−.097, .814) |
| U.S. region | ||||
| South | Ref | Ref | Ref | Ref |
| Midwest | −1.371 (−2.040, −.703) | −2.581d (−3.196, −1.966) | −3.101d (−3.786, −2.416) | .5624c (.199, .924) |
| West | −1.502 (−2.251, −.753) | −2.274d (−2.963, − 1.585) | −2.849d (−3.616, −2.083) | .130 (−.276, .536) |
| Northeast | −1.572 (−2.386, −.758) | −1.443d (−2.192, −.695) | −1.632d (−2.467, −.798) | −.065 (−.506, .377) |
| Average daily census | −.001 (−.002, .000) | −.0024c (−.003, −.001) | −.003d (−.004, −.002) | −.000 (−.001, .000) |
| Certification year | −.068d (−.104, −.032) | −.0464c (−.0794, −.013) | −.031 (−.068, .006) | −.036d (−.055, −.016) |
| Hospice practices | ||||
| Received the comprehensive assessment at admission | .0434c (.011, .075) | .0474c (.018, .076) | .041b (.008, .074) | .036d (.018, .053) |
| Visited near death | .152d (.126, .178) | .101d (.077, .126) | . 107d (.080, .134) | .074d (.060, .089) |
Estimated regression coefficients are listed followed by their 95% CI in parenthesis. Positive coefficients indicate that the independent variable is positively associated with the outcome reported by caregivers, and vice versa for negative coefficients. The strength of the association is determined by the value of the coefficient. Coefficients are marked with letter superscript only if the P-value is statistically significant (i.e. P < .05).
Government, combination government and non-profit, and any other possible ownership type.
P < .05.
P < .01.
P < .001.
CAHPS = Consumer Assessment of Healthcare Providers and Systems; CI = confidence interval.
Percentage of Caregivers who Reported That the Patient “Always” Received Help for Pain and Symptoms
Hospices with a higher proportion of patients with a diagnosis of dementia (P = .001), stroke (P < .001), or other conditions (P < .01) were associated with a lower percentage of caregivers who reported that they always received help for pain and symptoms (Table 3). The proportions of patients with diagnoses of cancer, heart disease, or respiratory disease were not associated with this outcome.
Percentage of Caregivers who Reported “Definitely” Having Received the Training They Needed to Care for Patient
Hospices with a higher proportion of patients with a dementia (P < .001), stroke (P < .001), or other conditions (P = .028) diagnosis were associated with a lower percentage of caregivers who reported having received the training they needed (Table 3). The proportions of patients with cancer, heart disease, or respiratory disease were not associated with this outcome.
Percentage of Caregivers who Received the “Right Amount” of Emotional and Spiritual Support
Lastly, hospices with a higher proportion of patients with a cancer (P < .001), stroke (P = .001), or other conditions (P = .001) diagnosis were associated with a lower percentage of caregivers who reported that the hospice offered the right amount of emotional and spiritual support (Table 3). The proportion of patients with dementia, heart disease, or respiratory disease was not associated with this outcome.
Discussion
Our study found associations between the composition of hospice patients’ terminal diagnosis and several important caregiver-reported quality measures. In particular, we found that hospices that cared for a larger percentage of patients with dementia or stroke were associated with lower quality scores in all 4 outcomes examined. These results highlight the potential differences in end-of-life care based on a patient’s terminal diagnosis and support the need for future work to understand these associations better, with an ultimate goal of improving hospice care for all patients and caregivers, regardless of diagnosis.
We found that hospices that cared for a higher percentage of patients with dementia were associated with lower caregiver-reported scores on the quality measures. We speculate that this finding could be due to clinical challenges that often occur when providing care to patients with dementia, who are often unable to communicate their needs, which may make it difficult to optimize care given the challenges of ascertaining their care preferences. In addition, a contributing factor that we were unable to study in our analysis was the setting (e.g., home, hospital, hospice inpatient) in which patients received hospice care. Prior studies have reported better quality of hospice care when it is delivered at home vs in nursing homes (NH) and assisted living facilities (ALF)16,17 and a greater proportion of dementia patients on hospice receive care in an NH and ALF when compared with other terminal diagnoses.17 This calls for further research looking into the care that is received in different hospice settings and how that may impact the differences we see in care for dementia patients at the end-of-life.
We also found that hospices with a greater percentage of stroke patients had lower caregiver-reported quality measures. While there are sparse data examining stroke patients and their experience with hospice care, we believe that some reasons for our findings overlap with reasons outlined for dementia patients. In particular, the trajectory of end-of-life care for stroke patients can be uncertain and unpredictable for hospice providers, making it difficult to optimize end-of-life care.18 And similar to dementia patients, a larger proportion of stroke patients receive hospice care in an NH and ALF when compared to other terminal diagnoses.17
A possible reason for our finding that hospices that cared for more dementia and stroke patients scored lower on patients receiving help for pain and symptoms might be due to the way this outcome is defined. Pain may not be as common of a symptom in these conditions in comparison to a patient with metastatic cancer, for example. Therefore, caregivers of stroke and dementia patients might have been less likely to report improvement in pain as compared to those of cancer patients. This can explain why Mitchell et al. found that overall hospice quality of care was similar among hospice patients with dementia, cancer, and other chronic diseases; in that study, caregivers were instructed to not answer unmet needs for pain control if their family member did not experience pain.8
One possible reason that NH or ALF hospice patients have lower caregiver-reported quality scores is the lack of effective communication between NH/ALF and hospice staff regarding goals of patient care, in addition to an absence of consensus regarding which of these parties should communicate with the caregivers.17,19,20 Therefore, caregivers of hospice patients living in NH/ALF can perceive the hospices as not providing adequate care and/or support to patients and their families. In contrast, for patients and caregivers receiving hospice care at home, there is a clear delineation of the role of the hospice team.
It is also important to take into consideration the heterogeneity of hospice care and the further implications of care when it is provided in different settings. In the home setting, families are responsible for providing the medical and physical care to the patient. The mindset and expectations of families who take on home hospice responsibilities may be different from families whose loved one receives hospice care in the NH/ALF setting. In addition, the priorities of the hospice staff in caring for the caregivers may also change based on setting. There may be less of a focus on education around symptom management or training for caregivers if a patient is residing in an NH/ALF since the responsibility of these duties may be assigned to trained staff members in the facility. Further research and information examining these dynamics will be important to help clarify the underlying mechanisms of our findings, but also to begin developing interventions that can improve patient and caregiver satisfaction with hospice care.
Lastly, while our objective was to examine caregiver-reported outcomes and hospice composition of patients by diagnosis, we did find an association between hospice profit status and caregiver-reported outcomes. Hospices that were non-profit or “other” were associated with a higher percentage of caregivers who rated a hospice a 9 or 10 when compared with for-profit hospices. This result is consistent with past research comparing quality of care delivered.21–25
Limitations
This study has limitations that warrant consideration. One limitation is that the survey response rate was 32%, which may introduce bias if there are differences between responding and non-responding hospice agencies. Data were not available to assess for differences between responders and non-responders. In addition, we only had access to caregiver-reported outcomes and not patient-reported outcomes, which is an acknowledged limitation in hospice research. Further, the dataset included variables at the hospice level and not at the patient or caregiver level, so no length of stay information or location of care was available. Moreover, research suggests that ethnic minorities face barriers to access high-quality palliative care. Therefore, a possible confounder in our study is the composition of patients by ethnicity that the dataset did not report.26 In addition, since the dataset did not provide information on individual survey responses, we did not have the ability to pool responses based on terminal diagnosis and compare different diagnoses directly with one another. Therefore, we were only able to examine outcomes based on the hospices’ composition of terminal diagnosis. Finally, the collection of hospice composition by diagnosis data in 2017 and survey results from 2018–2019 indicates that caregivers who rated the hospices were not necessarily the same caregivers who the diagnosis data were collected about. However, based on published data on US hospices, we see similar proportions of patient composition by diagnosis from 2016–2019.27 While there are trends over time (e.g., increased use by persons living with dementia since early 2000s), system-wide year on year numbers are fairly stable.3
Conclusion
Our study reveals an association between hospice patient composition by diagnosis and caregiver-reported hospice quality measures, independent of hospice characteristics and practice measures. These findings support efforts to identify mechanisms underlying these differences and to design and test strategies to ensure optimal outcomes for all hospice patients regardless of disease diagnosis.
Key Message.
Hospices that cared for a larger percentage of patients with dementia and stroke were associated with lower caregiver-reported quality outcomes. Understanding the underlying reasons for this is important to be able to develop interventions and/or implement policy changes aimed at improving care for all patients on hospice, regardless of diagnosis.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The author(s) disclosed receipt of the following financial support: Dr. Phongtankuel is supported by a grant from the National Institute of Aging (K76 AG059997) and Weill Cornell Medical College Clinical and Translational Science Center. Dr. Reid is supported by grants (K24AG053462, P30AGO22845, R01AG070055) from the National Institute on Aging. Dr. Luth is supported by a grant from the National Institute of Aging (AG065624).
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
References
- 1.Buckingham RW, Lupu D. A comparative study of hospice services in the United States. Am J Public Health. 1982;72(5): 455–463. doi: 10.2105/AJPH.72.5.455. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Aldridge MD, Canavan M, Cherlin E, Bradley EH. Has hospice use changed? 2000–2010 utilization patterns. Med Care. 2015; 53(1):95–101. doi: 10.1097/MLR.0000000000000256. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.National hospice and palliative care organization A (VA). NHPCO Facts and Figures 2020 Edition. Published 2020. Accessed September 2, 2021. https://www.nhpco.org/wp-content/uploads/NHPCO-Facts-Figures-2020-edition.pdf
- 4.De Vleminck A, Morrison RS, Meier DE, Aldridge MD. Hospice care for patients with dementia in the United States: A longitudinal cohort study. J Am Med Dir Assoc. 2018;19(7): 633–638. doi: 10.1016/j.jamda.2017.10.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Russell D, Diamond EL, Lauder B, et al. Frequency and risk factors for live discharge from hospice. J Am Geriatr Soc. 2017; 65(8):1726–1732. doi: 10.1111/jgs.14859. [DOI] [PubMed] [Google Scholar]
- 6.Luth EA, Russell DJ, Xu JC, et al. Survival in hospice patients with dementia: The effect of home hospice and nurse visits. J Am Geriatr Soc. 2021;69(6):1529–1538. doi: 10.1111/jgs.17066. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Campbell RW. Being discharged from hospice alive: The lived experience of patients and families. J Palliat Med. 2015;18(6): 495–499. doi: 10.1089/jpm.2014.0228. [DOI] [PubMed] [Google Scholar]
- 8.Mitchell SL, Kiely DK, Miller SC, Connor SR, Spence C, Teno JM. Hospice care for patients with dementia. J Pain Symptom Manage. 2007;34(1):7–16. doi: 10.1016/j.jpainsymman.2007.01.003. [DOI] [PubMed] [Google Scholar]
- 9.Harrison KL, Allison TA, Garrett SB, Thompson N, Sudore RL, Ritchie CS. Hospice staff perspectives on caring for people with dementia: A multisite, multistakeholder study. J Palliat Med. 2020;23(8):1013–1020. doi: 10.1089/jpm.2019.0565. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Clinical Practice Guidelines for Quality Palliative Care . Vol 79; 2004. doi: 10.1542/peds.2014-0046 [DOI] [PubMed] [Google Scholar]
- 11.Teno JM, Casey VA, Welch LC, Edgman-Levitan S. Patient-focused, family-centered end-of-life medical care: Views of the guidelines and bereaved family members. J Pain Symptom Manage. 2001;22(3):738–751. doi: 10.1016/S0885-3924(01)00335-9. [DOI] [PubMed] [Google Scholar]
- 12.Hospice care | Provider data catalog. Accessed August 20, 2021. https://data.cms.gov/provider-data/topics/hospice-care
- 13.U.S. Department of Commerce. Census Regions and Divisions of the United States. Published online 2010:1–2. https://www2.census.gov/geo/pdfs/maps-data/maps/reference/us_regdiv.pdf
- 14.Current measures | CMS. Accessed August 20, 2021. https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Hospice-Quality-Reporting/Current-Measures
- 15.Anhang Price R, Tolpadi A, Schlang D, et al. Characteristics of hospices providing high-quality care. J Palliat Med. 2020; 23(12):1639–1643. doi: 10.1089/jpm.2019.0505. [DOI] [PubMed] [Google Scholar]
- 16.Unroe KT, Stump TE, Effler S, Tu W, Callahan CM. Quality of hospice care at home versus in an assisted living facility or nursing home. J Am Geriatr Soc. 2018;66:687–692. doi: 10.1111/jgs.15260. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Quigley DD, Parast L, Haas A, Elliott MN, Teno JM, Price RA. Differences in caregiver reports of the quality of hospice care across settings. J Am Geriatr Soc. 2020;68(6):1218–1225. doi: 10.1111/JGS.16361. [DOI] [PubMed] [Google Scholar]
- 18.Steigleder T, Kollmar R, Ostgathe C. Palliative care for stroke patients and their families: Barriers for implementation. Front Neurol. 2019;10(March):1–8. doi: 10.3389/fneur.2019.00164. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Hwang D, Teno JM, Clark M, et al. Family perceptions of quality of hospice care in the nursing home. J Pain Symptom Manage. 2014;48(6):1100–1107. doi: 10.1016/j.jpainsymman.2014.04.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Shield RR, Wetle T, Teno J, Miller SC, Welch L. Physicians “missing in action”: Family perspectives on physician and staffing problems in end-of-life care in the nursing home. J Am Geriatr Soc. 2005;53(10):1651–1657. doi: 10.1111/j.1532-5415.2005.53505.x. [DOI] [PubMed] [Google Scholar]
- 21.Prsic E, Plotzke M, Christian TJ, Gozalo P, Teno JM. A national study of live hospice discharges between 2000 and 2012. J Palliat Med. 2016;19(9):987–990. doi: 10.1089/jpm.2015.0383. [DOI] [PubMed] [Google Scholar]
- 22.Stevenson DG, Grabowski DC, Keating NL, Huskamp HA. Effect of ownership on hospice service use: 2005–2011. J Am Geriatr Soc. 2016;64(5):1024–1031. doi: 10.1111/jgs.14093. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Cherlin EJ, Carlson MD, Herrin J, et al. Interdisciplinary staffing patterns: Do for-profit and nonprofit hospices differ? J Palliat Med. 2010;13(4):389–394. [DOI] [PubMed] [Google Scholar]
- 24.Barry CL, Carlson MDA, Thompson JW, et al. Caring for grieving family members: Results from a national hospice survey. Med Care. 2012;50(7):578–584. doi: 10.1097/MLR.0b013e318248661d. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Carlson MDA, Gallo WT, Bradley EH. Ownership status and patterns of care in hospice: Results from the national home and hospice care survey. Med Care. 2004;42(5):432–438. doi: 10.1097/01.mlr.0000124246.86156.54. [DOI] [PubMed] [Google Scholar]
- 26.Gardner DS, Doherty M, Bates G, Koplow A, Johnson S. Racial and ethnic disparities in palliative care: A systematic scoping review. Fam Soc. 2018;99(4):301–316. doi: 10.1177/1044389418809083. [DOI] [Google Scholar]
- 27.National Hospice and Palliative Care Organization. Hospice Facts and Figures Table of Contents; 2021. https://www.nhpco.org/wp-content/uploads/NHPCO-Facts-Figures-2021.pdf.