Live Discharge of Hospice Patients with Alzheimer’s Disease and Related Dementias: A Systematic Review

Abstract

Background:

Hospice is intended to promote the comfort and quality of life of dying patients and their families. When patients are discharged from hospice prior to death (i.e., experience a “live discharge”), care continuity is disrupted. This systematic review summarizes the growing body of evidence on live discharge among hospice patients with Alzheimer’s Disease and related dementias (ADRD), a clinical subpopulation that disproportionately experiences this often burdensome care transition.

Methods:

Researchers conducted a systematic review in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Reviewers searched AgeLine, APA PsycINFO (Ovid), CINAHL Plus with Full Text, ProQuest Dissertations & Theses Global, PubMed, Scopus, and Web of Science (Core Collection). Reviewers extracted data and synthesized findings from 9 records, which reported findings from 10 individual studies.

Results:

The reviewed studies, which were generally of high quality, consistently identified diagnosis of ADRD as a risk factor for live discharge from hospice. The relationship between race and live hospice discharge was less clear and likely dependent upon the type of discharge under investigation and other (e.g., systemic-level) factors. Research on patient and family experiences underscored the extent to which live hospice discharge can be distressing, confusing, and associated with numerous losses.

Conclusion:

Research specific to live discharge among ADRD patients and their families is limited. Synthesis across included studies points to the importance for future research to differentiate between types of live discharge—revocation versus decertification—as these are vastly different experiences in choice and circumstances.

Introduction

In the United States, costs associated with most hospice patients’ care are covered under the Medicare Part A Hospice Benefit (MHB), which requires that hospice enrollees have a life-limiting illness and corresponding prognosis of 6 months or less.^1,2 In 2019, U.S. hospice agencies served more than 1.6 million MHB beneficiaries, approximately 20.9% of whom had a primary diagnosis of Alzheimer’s Disease or a related dementia (ADRD).² Hospice patients with ADRD received more days of care on average than patients with any other primary diagnosis (143 days), and they were significantly more likely to be discharged from hospice alive, a transition of care referred to as a “live discharge.”^2–4

Live discharges can be initiated by patients/proxies or by hospice providers.³ Most patient-initiated discharges (or “revocations”) occur to enable patients to access disease-directed therapies or inpatient hospitalization, neither of which are covered by the MHB and therefore, a patient must revoke this benefit.⁵ In contrast, most hospice-initiated discharges occur when a patient’s condition stabilizes or improves and their prognosis is extended, rendering them ineligible for services.⁵ This is sometimes informally referred to as “graduation” from hospice; however, this framing has been criticized as misleadingly euphemistic, given that this care transition is often unwanted, unanticipated, and can result in significantly increased patient and family burden.^3,4 In 2019, 17.4% of U.S. hospice patients were discharged alive from hospice; over one-third of these discharges were attributed to patients’ extended prognoses.²

The higher proportion of live discharges among hospice patients with ADRD relative to other terminal conditions is largely explained by the challenging nature of prognostication in dementia care.⁶ For years, individuals with ADRD experience pronounced functional impairment with no clear patterns for predicting timing of death, differing significantly from patients with diseases such as cancer that typically have more abbreviated and predictable end-of-life trajectories.^7,8 As a result, it is not at all uncommon for patients with ADRD to satisfy hospice admission criteria, enroll into hospice, and then remain functionally stable or even appear to improve in response to an increased level of care. When this occurs, patients no longer meet the 6-month prognosis requirement and are discharged alive from hospice.⁴

Statement of Purpose

The significance of live discharge from hospice for individuals with ADRD combined with a growing body of studies related to this topic led our team of researchers to conduct the systematic review described herein. The overarching goal of the review was to synthesize the evidence on live discharge from hospice for patients with ADRD and their families. Four questions guided the review: (1) What is the prevalence of live discharge among hospice patients with ADRD? (2) What factors are associated with the live discharge of hospice patients with ADRD? (3) What are the experiences of patients with ADRD discharged alive from hospice and their families? and (4) What interventions related to hospice live discharge for patients with ADRD have been systematically evaluated? In answering these questions, we extend prior reviews⁹ by focusing specifically on ADRD as a clinical population of interest.

Methods

We designed our systematic review protocol in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.¹⁰ To identify reports of studies for potential inclusion in the review, we searched AgeLine, APA PsycINFO (Ovid), CINAHL Plus with Full Text, ProQuest Dissertations & Theses Global, PubMed, Scopus, and Web of Science (Core Collection), implementing a search strategy (Table 1) developed by our team’s research librarian. We then uploaded all identified reports into Rayyan¹¹ to facilitate duplicate record removal, blinded abstract screening by at least two independent reviewers, and final eligibility assessment of full-text reports (details provided in Figure 1). Reports were included in the review if they described original research, were published in English between January 1, 2012 and December 7, 2021 (the date of the search), and explicitly included or described the experiences of hospice patients with ADRD and/or their family caregivers. January 1, 2012 was chosen as the initial search start date as it was the beginning of the first full year after publication of the Office of Inspector General’s (OIG) report that focused on nursing home hospice enrollees as a key reason for high claims of Medicare beneficiaries, many of whom may not have met ongoing eligibility requirements for hospice.¹² The OIG report and subsequent results (e.g. hospice closures) greatly impacted hospice live discharge practices across the US. Editorials, conceptual or theoretical pieces, and other reviews were excluded, as were conference abstracts that were later published as research articles or included insufficient detail with regard to study results.¹³ All conflicts regarding record inclusion or exclusion were resolved via discussion among reviewers and arrival at consensus.

Table 1.

Search Queries by Database

Database	Search Query
PubMed.gov	(“Hospices”[mesh] OR “Hospice Care”[Mesh] OR hospice*) AND (“Live discharge” OR “live hospice discharge” OR “discharged alive” OR “extended prognosis” OR decertif* OR “Patient Discharge”[Mesh] OR stabiliz* OR terminat*) AND (alzheimer* OR “Alzheimer Disease”[Mesh] OR “Dementia”[Mesh] OR dementia*) AND (English[Language])
APA PsycInfo (1987-present) (OVID)	((hospice*).mp. or (hospices or “hospice care”).mh. or (hospice).sh.) AND ((“Live discharge” or “live hospice discharge” or “discharged alive” or “extended prognosis” or decertif* or stabiliz* or terminat*).mp. or patient discharge.mh.) AND ((alzheimer* or dementia*).mp or (“alzheimer’s disease” or dementia).sh. or (“alzheimer disease” or dementia).mh.) AND english.lg.
CINAHL Plus with Full Text (EBSCOHost)	((MH “Hospice Care+”) OR (MH “Hospices+”) OR (MH “Hospice Patients+”) OR hospice*) AND (“Live discharge” OR “live hospice discharge” OR “discharged alive” OR “extended prognosis” OR decertif* OR (MH “Patient Discharge+”) OR stabiliz* OR terminat*) AND (alzheimer* OR (MH “Alzheimer’s Disease+”) OR (MH “Dementia+”) OR dementia*) AND (LA “English”)
Scopus	TITLE-ABS-KEY(hospice*) AND TITLE-ABS-KEY(“Live discharge” OR “live hospice discharge” OR “discharged alive” OR “extended prognosis” OR decertif* OR “Patient Discharge” OR stabiliz* OR terminat*) AND TITLE-ABS-KEY(alzheimer* OR dementia*) AND LANGUAGE(English)
Web of Science Core Collection	TS=hospice* AND TS=(“Live discharge” OR “live hospice discharge” OR “discharged alive” OR “extended prognosis” OR decertif* OR “Patient Discharge” OR stabiliz* OR terminat*) AND TS=(alzheimer* OR dementia*) AND LA=(English)
AgeLine (EBSCOHost)	hospice* AND “Live discharge” OR “live hospice discharge” OR “discharged alive” OR “extended prognosis” OR decertif* OR “Patient Discharge” OR stabiliz* OR terminat* AND alzheimer* OR dementia*
ProQuest Dissertations & Theses Global	noft(hospice*) AND noft(“Live discharge” OR “live hospice discharge” OR “discharged alive” OR “extended prognosis” OR “patient discharge” OR stabiliz* OR terminat* OR decertif*) AND (alzheimer* OR dementia*)

¹Initial search strategy developed in PubMed and adapted to additional databases

²Quotation marks must be sans serif/straight lines to work in APA PsycInfo (OVID)

³Must use advanced search for search string to work

⁴Search included specialty-specific metadata fields by selecting “Search in: All Databases” and “Collections: Web of Science Core Collection”. This yields more results than selecting the default “Search in: Web of Science Core Collection.”

⁵Language limiter is not an option in AgeLine

⁶“English” selected as language limiter before searching

Figure 1.

PRISMA Flow Diagram

*One of the reports included in the review corresponded to a “three paper” dissertation. Two of the 3 studies contained in the dissertation were determined to meet the review inclusion criteria. As a result, 10 studies described in 9 reports were ultimately included in the review.

Critical Appraisal of Study Quality

In light of our decision to review quantitative, qualitative, and mixed methods studies, we chose a modified version of the Mixed Methods Appraisal Tool¹⁴ (MMAT; Table 2) to guide our critical appraisal of the included studies’ quality. Each study’s quality was critically appraised by at least two independent reviews, who selected among the following responses for each applicable quality criterion: Yes, No, Unclear, or Not Applicable. Discrepancies in quality ratings were discussed among reviewers and resolved via consensus and, as needed, additional review by team members with relevant methodological expertise.

Table 2.

Modified Mixed Methods Appraisal Tool (MMAT) Methodological Quality Criteria

Study design category	Methodological quality criteria1
1. Quantitative	1.1. Is the sampling strategy relevant to address the research question?
	1.2. Is the sample representative of the target population?
	1.3. Are the measurements appropriate?
	1.4. Is the risk of nonresponse bias low/are there complete outcome data?
	1.5. Is the statistical analysis appropriate to answer the research question?
2. Qualitative	2.1. Is the qualitative approach appropriate to answer the research question?
	2.2. Are the data collection methods adequate to address the research question?
	2.3. Are the findings adequately derived from the data?
	2.4. Is the interpretation of results sufficiently substantiated by data?
	2.5. Is there coherence between data sources, collection, analysis, and interpretation?

¹Mixed methods criteria (not shown) were determined to be inapplicable to the present review.

Data Extraction

At least two reviewers independently extracted data from each of the included studies using a standardized form created specifically for this review. The form directed reviewers to extract the following data: full article citation, study aim(s) and/or research question(s), sample size and description, relevance to ADRD, study design and research methods, and key findings pertinent to our specific review questions. Discrepancies in extracted data were discussed among reviewers and resolved via group discussion and re-reading of the published studies.

Results

Nine reports were ultimately included in the review. One of the reports¹⁵ corresponded to a so-called “three-paper” dissertation (i.e., a dissertation comprising three distinct yet related studies¹⁶), and two of the three studies in the dissertation met our inclusion criteria, resulting in a total of 10 unique studies. Dolin,¹⁵ Luth,^17,18 and Wladkowski^19,20 each served as sole or lead author of two studies. A summary of the reviewed studies is provided in Table 3.

Table 3.

Summary of Reviewed Studies

Author(s), year of publication	Study aim(s) and/or research question(s)	Sample size and description	Relevance to ADRD	Study design and research methods	Key findings
Clayton et al., 2021	Study aims: (1) To examine live hospice discharge prevalence, and (2) to describe the experiences of patients’ families and hospice staff members related to live discharge.	Quantitative sample: 5,648 hospice admissions (795 live discharges) between 2013 and 2017 at a large non-profit hospice agency in the western United States. Qualitative sample: 11 (5 family caregivers, 3 hospice aides, 2 hospice nurses, 1 hospice administrator) from the same agency.	Descriptive statistics were calculated regarding prevalence of live discharge among hospice patients with ADRD. Hospice staff members discussed live discharge of patients with ADRD.	*Multiple methods: (1) quantitative evaluation of hospice admissions and discharge administrative records from a large non-profit hospice in the Western United States to calculate discharge rates, and (2) descriptive qualitative analysis of interview and focus group data to explore the experiences of patients’ families and staff members.	Dementia was the most prevalent live discharge diagnosis; 19% of patients with a dementia admitting diagnosis were ultimately discharged alive. Live discharges were described as distressing. Staff members understood regulatory pressures, but felt they were “abandoning” patients and families. Family caregivers were troubled by the uncertainty and burden added when hospice services ended.
Dolin, 2017a	Study aim: To ascertain provider perspectives on key drivers of live discharge from hospice.	18 U.S. hospice administrators representing 14 different hospices in 2016.	Interviewees specifically discussed patients with ADRD when describing prognostication challenges and reasons for increased CMS oversight.	Qualitative; template analysis of individual interview data.	Administrators described four key drivers of hospice live discharge: 1) difficulty estimating patient prognosis on admission, 2) increased CMS oversight, 3) rising market competition, and hospice business practices, and 4) challenges with inpatient contracting.
Dolin, 2017b	Research question: Do live discharge rates increase as hospices approach their Medicare aggregate payment caps?	2,026,456 Medicare beneficiaries discharged from hospice from 2012–2013 (273,665 live discharges, 145,475 of which were patient/proxy revocations and 116,963 of which were provider-initiated).	Percentage of total sample, deaths, and live discharges of hospice patients with ADRD were reported.	Quantitative; secondary analysis of Medicare hospice claims data (national sample).	Patients with ADRD accounted for 16.8% of all hospice discharges: 16.2% of deaths and 20.7% of live discharges. Results showed a positive, statistically significant relationship between a hospice’s risk of exceeding its aggregate payment cap and the likelihood that an enrolled patient would be discharged alive.
Johnson et al., 2012	Study aims: (1) To examine the characteristics of hospice enrollees with ADRD who were discharged alive because their condition stabilized or improved, and (2) to identify predictors of these patients’ deaths in the year after discharge.	24,111 hospice enrollees age 65 and older with a primary admission diagnosis of ADRD who died or were discharged alive after receiving services from a large, national, for-profit hospice provider between 1999 and 2003.	Analyses specifically examined live discharge among hospice patients with ADRD.	Quantitative; cross-sectional analysis of clinical and administrative data from a national for-profit hospice; a sub-analysis analyzed time to death for patients discharged alive.	Approximately 5% of hospice enrollees with ADRD were discharged alive because their condition stabilized or improved. Those discharged alive were slightly younger, more likely to be female, more likely to have an admission level of care of routine home care in a private residence, and more likely to have a length of stay exceeding 180 days than those who died. After controlling for demographic and hospice use variables, those discharged alive because of stabilization or improvement were less likely than those who died to be aged 85 and older, less likely to be AA than white, and less likely to have an admission level of care of routine home care in a nursing home, impatient care, or continuous care than routine home care in a private residence. The single greatest predictor of being discharged alive due to stabilization or improvement was length of stay. Of 303 patients with ADRD discharged alive, about 75% were still alive a year later; none of the studied demographic variables were associated with death after live discharge.
Luth et al., 2020	Study aims: (1) To identify demographic and health correlates of live discharge due to condition stabilization in hospice patients with ADRD, and (2) explore how hospice service delivery factors relate to this live discharge type for patients with ADRD.	2,629 hospice patients age 65 and older with ADRD who were receiving services from a large, non-profit agency located in New York City from 2013–2017. Patients with a length of stay shorter than 7 days or greater than 180 days were excluded.	Analyses specifically examined live discharge among hospice patients with ADRD.	Quantitative; retrospective cohort study of electronic health records from a large non-profit hospice in New York City.	7% of patients with ADRD experienced a live discharge due to condition stabilization. Hispanic patients represented a larger proportion of the patients discharged alive due to stabilization than deceased. Patients discharge alive due to stabilization were referred from hospitals less often and received home hospice more often than deceased patients. On average, patients discharged alive had a longer length of service and received more nurse visits than deceased patients. When controlling for patient demographic and clinical factors and hospice service characteristics, compared with white hospice patients with dementia, African American and Hispanic hospice patients with dementia experienced increased risk of live discharge due to stabilized condition. Higher functional status, receipt of home hospice care, longer length of service were also associated with live discharge. Number of nurses’ visits and days from the last nurse visit to hospice discharge date were associated with live discharge; however, this relationship was nonlinear.
Luth et al., 2021	Study aim: To examine how home hospice and nurse visit frequency relate to dying in hospice within the Medicare-intended 6-month period.	3,837 patients age 65 and older with ADRD who received hospice services from a large, non-profit hospice agency located in New York City from 2013–2017. Patients with a length of stay of fewer than 7 days in hospice were excluded.	Analyses specifically examined death within 6 months of hospice admission among patients with ADRD.	Quantitative; retrospective cohort study of electronic medical records from a large non-profit hospice in New York City.	Approximately 5% of hospice patients with ADRD were discharged due to condition stabilization. Patients who receive home hospice care were at increased risk for live discharge compared to patients who received care in nursing homes. Frequency of nurse visits was inversely associated with live discharge.
Russell et al., 2017	Study aim: To report frequencies and associated risk factors for four distinct causes of live discharge from hospice: acute hospitalization, elective revocation to resume disease-directed treatments with the goal of life prolongation, disqualification, and transferring.	9,190 adult home hospice patients served by a large, urban, non-profit organization in New York City between 2013 and 2015.	Primary patient diagnosis (e.g., ADRD) was included as an analytic variable.	Quantitative; retrospective cohort study of electronic health records from a non-profit hospice in New York City.	Relative to patients with a primary diagnosis of cancer, the odds of being discharged due to acute hospitalization or disqualification due to loss of eligibility were significantly greater among those with a primary diagnosis of dementia; however, patients with a primary diagnosis of dementia had lower odds of transferring to another hospice when compared to patients with a primary diagnosis of cancer.
Teno et al., 2015	Study aim: To characterize hospices with high rates of problematic patterns of live discharges including burdensome transitions (i.e., hospice discharge-hospitalization-hospice readmission), live discharge in the first seven days of hospice service, and live discharge after 180 days of hospice care.	3,028 hospice programs that collectively discharged 996,208 Medicare fee-for-service beneficiaries in 2010.	Provider rates of medical diagnoses (including ADRD) were reported and included in analyses.	Quantitative; secondary analysis of Medicare hospice claims data (national sample).	For-profit hospice providers, especially those without a chain affiliation, had the highest likelihood of having problematic live discharge patterns.
Wladkowski, 2016	Study aim: To explore the grief experiences of caregivers of adults with ADRD who experienced a live discharge from hospice.	24 adult primary caregivers of an adult with ADRD who was discharged alive from hospice.	All participants had experience caring for an adult with ADRD who was discharged alive from hospice.	Qualitative; thematic analysis of individual interview data.	Caregivers of hospice patients with ADRD reside on a continuum of grief that is complicated by live discharge, which requires them to reconcile co-existing realities of their care recipient being both “terminal” and “not dying fast enough.”
Wladkowski, 2017	Study aim: To explore the experiences of family caregivers of adults with ADRD who experienced a live discharge from hospice.	24 adult primary family caregivers of an adult with ADRD who was discharged alive from hospice.	All participants had experience caring for an adult with ADRD who was discharged alive from hospice.	Qualitative; thematic analysis of individual interview data.	Caregivers often experience live discharge as an abrupt end that results in multiple losses, including loss of services from the hospice team, loss of coverage for equipment and supplies, and loss of the flexibility to decline medical care or choose to have care provided on-site.

Note. ADRD = Alzheimer’s Disease and related dementias; CMS = Centers for Medicare and Medicaid Services.

^*Reviewers characterized this study as employing multiple (versus mixed) methods, as quantitative and qualitative methods addressed distinct study aims.

Critical Appraisal of Study Quality

As shown in Table 4, the reviewed studies were generally of high quality. Nearly all (33/35) of the methodological quality criteria were satisfied for studies employing quantitative methods, and 100% (20/20) of the methodological quality criteria were satisfied for studies employing qualitative methods. Methodological quality criteria specific to mixed methods studies were not applied, as reviewers determined that the qualitative and quantitative components of the only mixed methods study²¹ included in the review addressed functionally distinct research questions.

Table 4.

Methodological Quality of Reviewed Studies

Author(s), publication year	1.1	1.2	1.3	1.4	1.5	2.1	2.2	2.3	2.4	2.5
Clayton et al., 2021	Y	Y	U	Y	U	Y	Y	Y	Y	Y
Dolin, 2017a	N/A	N/A	N/A	N/A	N/A	Y	Y	Y	Y	Y
Dolin, 2017b	Y	Y	Y	Y	Y	N/A	N/A	N/A	N/A	N/A
Johnson et al., 2012	Y	Y	Y	Y	Y	N/A	N/A	N/A	N/A	N/A
Luth et al., 2020	Y	Y	Y	Y	Y	N/A	N/A	N/A	N/A	N/A
Luth et al., 2021	Y	Y	Y	Y	Y	N/A	N/A	N/A	N/A	N/A
Russell et al., 2017	Y	Y	Y	Y	Y	N/A	N/A	N/A	N/A	N/A
Teno et al., 2015	Y	Y	Y	Y	Y	N/A	N/A	N/A	N/A	N/A
Wladkowski, 2016	N/A	N/A	N/A	N/A	N/A	Y	Y	Y	Y	Y
Wladkowski, 2017	N/A	N/A	N/A	N/A	N/A	Y	Y	Y	Y	Y

Note. Y = Yes. N = No. U = Unclear. = Not Applicable.

Prevalence and Factors Associated with Live Discharge among Hospice Patients with ADRD

The vast majority of reviewed studies addressed the prevalence of and/or factors associated with live discharge among hospice patients with ADRD. Most^{15,17,18,21–24} employed quantitative methods and involved secondary data analyses. One¹⁵ employed qualitative methods and involved primary data collection and analysis.

Synthesizing the available evidence on these topics was made challenging by the diversity of date ranges, geographic regions, and samples/populations (e.g., only patients with ADRD vs. patients with ADRD in addition to patients with other diagnoses) studied in the reviewed research; however, some clear patterns emerged. First, while death remains a much more common reason for hospice exit than live discharge, a diagnosis of ADRD is consistently found to increase the risk of live discharge, and hospice patients with ADRD typically comprise a larger proportion of live discharges than do patients with other diagnoses.^15,21,22 Receipt of hospice care in a home setting (i.e., private residence) rather than a nursing home or other facility is also fairly consistently identified as a risk factor for live discharge, ^17,18,24 though little to no data exist in these studies to explain this finding beyond higher functionality at time of enrollment for patients at home.¹⁷

Less clear is the influence of race and ethnicity on hospice live discharge. While Johnson and colleagues²⁴ found Black race to decrease a patient’s likelihood of live discharge relative to white race, Luth and colleagues¹⁷ found Black patients and Hispanic patients to be at increased risk of live discharge. These contrasting findings may be partially due to the fact that the two studies differed substantially with regard to dates, patients, and provider types examined. In addition, the reviewed research suggests that race and ethnicity may be differentially associated with different types of live discharge. For example, Russell and colleagues²² found minoritized race and ethnicity to be specifically associated with live discharge due to acute hospitalization and transfers or service moves, but not decertification, or when someone is determined to be “no longer terminally ill” and they are not recertified. Moreover, Teno and colleagues²³ highlighted associations between race and other variables (e.g., profit status²³) that are known influencers of live discharge. Finally, Dolin¹⁵ considered system-level data, finding that hospices with higher rates of live discharge also had much larger proportions of Black enrollees compared to hospices with lower rates.

Although each of the studies^15,23 that focused on the role of organizational factors in hospice live discharge examined different research questions, taken together, the results of this body of evidence underscore the importance of policies—particularly with regard to reimbursement for services—in understanding this care transition. Payment caps, increased (and increasingly aggressive) competition, and intensified federal oversight were qualitatively identified as drivers of live hospice discharge,¹⁵ and for-profit status was shown to be quantitatively associated problematic patterns of live discharge (i.e., hospice discharge-hospitalization-hospice readmission, live discharge in the first seven days of hospice service, and live discharge after 180 days of hospice care).²³

Experiences of Patients with ADRD Discharged Alive from Hospice and Their Families and Interventions to Support Them

Four studies^19–21,24 described the experiences of patients with ADRD and their families upon live discharge from hospice, three qualitatively^19–21 and one quantitatively.²⁴ As opposed to studies on live discharge prevalence and associated factors, studies focused on this topic were readily synthesized and illustrated a fairly consistent picture of the impact of live discharge on patients’ and families’ lives. Generally speaking, live discharges were described as distressing, confusing, and associated with numerous losses—both practical and relational—for patients and their families.^19–21 Clayton and colleagues²¹ studied hospice staff members in addition to patients’ families, finding live discharges to be upsetting for both groups of individuals. Johnson and colleagues²⁴ analyzed data on patient experiences after live discharge, finding that approximately 75% of the patients in their study who were discharged from hospice due to condition stabilization were still alive one year later. Wladkowski’s study¹⁹ reported contradictory feelings, including caregivers may have felt prepared for the live discharge, but unprepared for a change in terminal prognosis, yet this provided unanticipated additional time with the hospice patient. Of note is the fact that the experiences reflected in these studies were primarily those of patients and families experiencing live discharge due to stabilization or improvement of the patient’s condition. The experiences of patients and families who themselves initiated hospice discharge are largely absent from the reviewed research.

None of the studies included in the review involved the systematic evaluation of interventions (operationally defined as “purposively-implemented change strategies”^20,25) intended to affect the live discharge process or the individuals (e.g., patients, families, hospice providers) involved or impacted.

Discussion

With higher rates of live discharge occurring among patients with ADRD compared with other diagnoses,^15,21,22 a closer study of factors and experiences of this population is warranted, to which this review contributes, expanding on Wu and Volker’s⁹ previous review. Research specific to live discharge among ADRD patients and their families is limited (6 of the 10 articles included in this review focused solely on patients with ADRD or their caregivers^17–21,24), though synthesized findings point to important areas for further exploration. Similar to Wu and Volker’s review,⁹ patient (ADRD diagnosis, location of care, race/ethnicity) and organizational (payment caps, federal oversight, profit status) factors were associated with determining likelihood for live discharges to occur. However, there were some discrepancies (in race and ethnicity as discussed above) or limited data providing additional explanation of findings (e.g. care at home versus facilities). Synthesis across included studies points to the importance of differentiating between types of live discharge—revocation versus decertification—as these are vastly different experiences in choice and circumstances, each needing more in-depth study.

Only four of the included studies focused solely on decertification among ADRD patients,^17,19,20,24 with a few separating rates^15,18 and/or analysis related to each type,²² with remaining studies not differentiating between types of live discharge.^15,21,23 Studying live discharge inclusive of both revocation and decertification together has been an important starting point in an area with limited research, but future research must begin teasing out the different reasons and outcomes associated with each area. Additionally, there is currently no research examining the experience of hospice revocation, an important consideration among this population as more patients with ADRD experienced live discharge due to hospitalization (7%) than condition stabilization (5%), with a smaller percentage choosing revocation without immediate hospitalization (2%).¹⁸ Understanding motivations, experiences, and outcomes for those who choose to revoke hospice can assist in identifying areas for improvement during this burdensome transition.³

Findings related to organizational and systemic factors provide additional support for differentiating between discharge types in future research. While for-profit status of hospice companies is associated with problematic patterns of live discharge connected to both revocation (hospice discharge-hospital admission-hospice readmission or live discharge within the first 7 days) and decertification (live discharges after long lengths of stay, or 180+ days), differentiating between them might point to targeted approaches for change. Impacts of policy on live discharge practices, such as payment caps and federal oversight, seem particularly related to hospice decertification and longer-term hospice patients.²⁶ Separately, policy limitations on concurrent care, or curative and palliative care provided at the same time (hospice policy requires that patients forego curative care), encourages hospices to quickly seek out revocations from any hospice patient that goes to the Emergency Department or hospital for care. Authors of this review purposely sought out articles published post-2012 to capture current practices of live discharge following the 2011 OIG report¹² and intensive scrutiny into long-term hospice patients. Additional relevant hospice policies over the past 10 years include: elimination of failure-to-thrive and debility as qualifying hospice diagnoses enacted in 2013; the 2014 Improving Medicare Post-Acute Care Transformation (IMPACT) Act, which “systemized audits of long hospice stays;” and the January 2016 payment change creating two-tiers of payment based on length-of-stay (decreasing the reimbursement rate after 60 days).²⁷

Recently released work from Gianattasio and colleagues²⁷ that examined live discharge rates related to the latter two policy changes in 2014 and 2016 demonstrated that, while rates were declining prior to policy implementation, trends show patterns of stabilization post-policy enactment. No research exists that examines rates of decline between 2008 to mid-2013 (as shown in their analysis), but it could be related to the intensified federal oversight discussed by hospice administrators in Dolin’s¹⁵ work. Results from Gianattasio and colleagues²⁷ also demonstrated that trends were similar between patients with lung cancer and ADRD and that trends were driven mostly by for-profit hospices. The latter supports results from both this and Wu and Volker’s⁹ reviews. Regardless of the reasons for the declining trend prior to the 2014 and 2016 policy changes, findings from this review suggest that examining differences in live discharge trends by types of discharge is potentially an important missing variable to this recently published study.

In addition to differentiating between types of live discharge, more research is needed on the longitudinal experience of live discharge and on intervention studies to help patients and families who have experienced or are experiencing a live discharge. Only one study in this review²⁴ followed patients for a long period of time after the discharge, and it is the earliest time period of live discharges included. No studies focused on interventions, though findings show agreement on the distressing impact of live discharges on patients and families. While one study¹⁹ reported on contradictory feelings that a live discharge can present for caregivers, it is unclear if the “benefit” of additional time outweighs the distress associated with the live discharge. Wladkowski and Wallace³ refer to live discharge as the “forgotten misdiagnosed care transition” as it has received such little investigation as a medical care transition, is often misrepresented as a celebratory “graduation.”^4,28 Further, there is no guiding policy with clear guidelines or required discharge planning, unlike those in hospitals, skilled nursing facilities, and home health agencies. Finally, more research is needed on race and ethnicity, both as independent predictors of live discharge and in combination with other possible or known predictors.

Limitations

While findings point to important areas for future research on live discharge, some limitations exist. First, because several of the studies did not differentiate between the types of live discharge in their analysis, our synthesis is unable to account for some differences that may exist in comparing “apples to oranges” (or data specific to decertification and data including both decertification and revocation). Additionally, because we chose to focus solely on ADRD in this review, our decision to exclude any study that did not explicitly state an inclusion of ADRD patients means that some studies in which ADRD patients were included may have been missed. We felt it was important to include studies examining live discharge across diagnoses, as key findings could provide important differences to patients with ADRD. This review adds to current literature by synthesizing live discharge research related to patients with ADRD, resulting in important discussion for differentiating between types of live discharge in future research.