Interventions for behavioral health comorbidities in the hospice setting: a scoping review

Introduction

Background

Palliative care and hospice aim to deliver holistic care focused on improving the quality of life for individuals with serious illnesses. To support this goal, hospice and palliative medicine services seek to provide care across multiple domains, including physical, spiritual, psychological, and social (1). While palliative care is delivered across serious illness trajectories, hospice is a specific model of palliative care focused on end-of-life care. In the United States (U.S.), hospice utilizes a capitated care model to provide comprehensive person-focused care in the last six months of life (2). Hospice is a highly significant model of palliative care delivery from a public health perspective; in the U.S., approximately half of all Medicare decedents receive hospice services (3). Despite its significance and reach, hospice may be underrepresented in clinical serious illness care research because of challenges conducting research in hospice settings (4).

Patients living with serious illness have significant behavioral health (BH) burdens [defined as mental health conditions and substance use disorders (5)]. A large minority (~40%) of individuals in palliative care settings (including hospice) experience mood and anxiety symptoms, and more than half of all individuals in end-of-life care settings experience delirium (6-12). Over the past several years, there has been increasing recognition that palliative care fails to fully address the significant psychological burden imposed by serious illness (13-15). In response to this recognition, a growing body of scholarship has addressed the epidemiology (7,8), impact (16), assessment (17-19), and management (20,21) of BH conditions in the palliative care setting.

Rationale and knowledge gap

Recent scholarship on optimizing BH components of palliative care has largely been situated in ambulatory and inpatient palliative care settings. However, there are key differences between hospice and other palliative care settings. From a patient perspective, hospice patients must have a prognosis of six months or less for eligibility, while general palliative care includes patients living with serious illnesses but not approaching end-of-life (22). Structurally, many palliative care programs operate in a consultative role in a medical system with access to specialist clinicians, including mental health referral resources. In contrast, hospice provides comprehensive care. Furthermore, hospice utilizes a unique delivery system and interdisciplinary team structure (which in the U.S. must include physicians, chaplains, social workers, and nurses). In the U.S., hospice is often oriented around delivering care in the home or nursing home setting. As such, interventions targeting BH conditions in U.S. hospice care generally must be deployable in the home setting. In other regions such as the United Kingdom (UK), hospice may be delivered in specialized medical settings such as residential hospices, which may also have unique BH workforce limitations relative to general medical settings (23). In the U.S., because of the capitated payment model, increasing the hospice workforce to include BH experts may not be financially feasible; therefore, ideal interventions should be deliverable by the core hospice interdisciplinary team. Finally, because the median length of hospice utilization is often short [48 days in the UK (24), 18 days in the U.S. (3), and 20 days in international models of home-based palliative care (25)], interventions must be feasibly delivered over a relatively brief period, which may be incongruent with many commonly utilized mental health treatments (e.g., psychotherapy, serotonergic antidepressants).

While existing literature reviews synthesize epidemiologic and intervention data on BH research in general palliative care settings (14-16,21), few studies specifically focused on the hospice setting. Given the significance of hospice as a widely disseminated, highly utilized model of palliative care delivery with distinct structural and clinical delivery features, this represents a substantial gap in the literature.

Objective

We sought to review BH intervention studies in the hospice setting. We define BH interventions as interventions specifically targeting mental health and substance use disorders or psychiatric symptoms (e.g., depression, anxiety) as their primary therapeutic target. Our decision to conduct a scoping rather than a systematic review was driven by our primary objective: to map existing literature and to characterize the state of BH intervention research in the hospice setting rather than to delineate clinical guidelines or recommendations. In doing so, we sought to identify gaps and opportunities to develop research and clinical paradigms to address the BH needs of patients receiving hospice care. We present this article in accordance with the PRISMA-ScR reporting checklist (26) (available at https://apm.amegroups.com/article/view/10.21037/apm-23-508/rc).

Methods

We conducted a modified scoping review utilizing Arksey and O’Malley’s scoping review framework (27) with enhancements to leverage team-based science, including the use of expert consultants (authors M.C.R. and V.P.) and representation of both clinical and research perspectives (28). Arksey and O’Malley outline a six-step process by which to conduct scooping reviews. Step 1: the research question was identified by the senior author (a clinician-investigator with expertise in BH and hospice and palliative medicine) in collaboration with the study team and research librarian (see the Objective subsection in the Introduction). Step 2: a research librarian (co-author D.W.) refined the search strategy based on the research question (see the Literature search subsection in the Methods and Appendix 1). Step 3: study selection occurred based on study protocol (see the Inclusion and exclusion criteria subsection and the Study selection and data extraction subsection, both in the Methods). Step 4: data were extracted and charted (see the Study selection and data extraction subsection in the Methods). Step 5: data were synthesized and presented (see the Study selection and data extraction subsection and the Results section). Step 6: we also utilized the optional expert consultation step noted above. In addition to emphasizing team-based science, we modified the typical scoping review methodology by conducting this study as a “nested” review; our initial search and screening processes were for general BH studies in hospice. At the full-text review stage, we delineated content areas (intervention versus epidemiology/outcomes/assessment) to generate two interrelated reviews from our initial search and screening. We conducted a scoping rather than a systematic review because scoping reviews are preferred when the foci entail identifying knowledge gaps and characterizing existing evidence in a field (26). We did not register a review protocol a priori.

Literature search

A comprehensive search of four databases (PubMed/Medline, Embase, Cochrane, and Scopus) was performed by a research librarian (co-author D.W.) with the strategy developed in collaboration with senior author D.S. (see Appendix 1). The search strategy included relevant vocabulary for concepts including hospice care and mental/BH (see Appendix 1). To be as expansive as possible and limit selection bias, our initial search was performed without limiting articles by publication date, language, or article type. The initial search was conducted on 11/9/2021 and then updated on 5/25/2023.

Inclusion and exclusion criteria

We used the following eligibility criteria for included studies:

• Study design: we included clinical trials (including both controlled and uncontrolled clinical trials) as well as observational and pseudo-experimental studies of interventions.
• Population: we included studies of adult hospice patients either exclusively or as at least 33% of the study population. For studies conducted outside the U.S., we also included studies in which at least 33% of patients received care at a hospice-equivalent end-of-life care paradigm (e.g., a care setting providing care focused on quality of life to patients with a prognosis of six months or less). We set this threshold based on our goal that an intervention had been substantially evaluated in hospice patients and our concern that some studies may have theoretically included hospice patients but not successfully recruited a substantial portion of such patients relative to patients in other settings.
• Intervention: we included studies of BH interventions (psychotherapy or other behavioral interventions, psychopharmacology) and studies of complementary and alternative interventions if a BH outcome was measured as a primary or secondary outcome. We used the National Center for Complementary and Integrative Health classification recommendations to define complementary and alternative interventions (29).
• Outcomes: we included studies that included a primary or secondary BH outcome (e.g., depression symptoms, anxiety symptoms, psychological distress, agitation/confusion). We excluded studies focused solely on caregiver outcomes.
• Publication status and language: we included English-language conference abstracts and full-text articles published in peer-reviewed journals.

Study selection and data extraction

Retrieved studies were screened for inclusion using Covidence (30), a web-based literature review platform. Two independent reviewers reviewed the titles and abstracts against the initial inclusion/exclusion criteria. Conflicts were resolved by D.S., the senior investigator, and a physician trained in both psychiatry and palliative medicine. All empirical studies involving BH in the hospice setting were eligible for inclusion at this stage. Studies screened for full-text review were initially designated as intervention or non-intervention studies. Given that this study included only patient-focused interventions, studies focused on caregivers were excluded from the present review. Two independent study team members evaluated full-text intervention studies with discrepancies resolved by senior author D.S..

Data extraction

Data extraction was conducted using a templated tool integrated into the Covidence platform (see Appendix 2). Primary author L.B. extracted data from each article and subsequently verified by senior author D.S. Extracted data included: year(s) of data collection, study location, study design, analytical sample size, sample characteristics (including percent of study population in hospice and percent of study population with oncologic disease), hospice setting (e.g., inpatient or home), intervention details (type of intervention, description, length, interventionist), category of outcome (BH, quality of life, other), outcome measures used, and study results.

Critical appraisal

We elected not to appraise the quality of the literature, given our primary objective of mapping the literature rather than synthesizing evidence about specific interventions.

Data synthesis

We synthesized evidence along two domains. In mapping the literature, we collected data on both study characteristics (Tables 1-3) and the nature of BH interventions and results (Table 4). We present study design elements in aggregate and categorize BH interventions across the studies as behavioral, pharmacologic, complementary/alternative, or other.

While analyzing data, we recognized that heterogeneity in the use of the term hospice between health systems presented an unexpected challenge. While some health systems (e.g., those of the U.S. and the UK) distinguish the hospice model of care from other palliative care services, this distinction is not universally recognized. Furthermore, the definition of hospice varies across health systems. To avoid excluding potentially relevant studies but also to highlight research in hospice as a unique care delivery model, we present data for all identified articles and also present subgroup analyses of studies conducted in the UK and Ireland (68-70), Australia (71), New Zealand (72), the U.S. (73), Canada (74), Ireland, and Poland (75). These regions were included because they utilize hospice models consisting of distinct structures of care, at least some of which are delivered in the community/home setting and targeted towards individuals at the end of life. However, even among these systems, there is significant variability in how care is delivered at home versus facilities, regulations regarding concurrent disease-modifying treatments, distinct hospice providers, and payment models. We subsequently refer to these as the hospice model subgroup.

We excluded other regions from the subgroup analysis if the model of hospice appeared to consist predominantly of inpatient palliative care wards/units [e.g., China (76), Japan (77,78)], if there seemed to be no distinction between hospice and palliative care services [e.g., India (79), Bosnia (80)], or if we were unable to clarify the model of care within a given health system.

Results

Data extraction

The initial result yielded 12,062 results at the screening stage (Figure 1). Upon removing duplicates, we conducted title and abstract screening on 7,672 citations. Three hundred eighty-two studies were selected for full-text review. On categorization, 92 full texts were identified as intervention studies and evaluated for inclusion in this study. Following full-text review, 37 studies were included and underwent data extraction (Table 1).

Figure 1 PRISMA-ScR diagram.

Study characteristics

Studies included were from populations globally, most commonly the UK (29.7%) (36,39,41,42,45,46,49-51,55,60,61,65) and the U.S. (24.3%) (37,38,44,48,52,56,63,64,66). Twenty-eight studies were from hospice model subgroup regions (U.S., UK, Canada, Australia, New Zealand, Ireland, and Poland). Most studies were conducted in inpatient hospices (35%) or multiple hospice types (24%), and over two-thirds of studies enrolled primarily cancer patients (Table 2). Among hospice model subgroup studies, fewer (25%) of studies were predominantly inpatient. In contrast, most studies conducted in health systems that deliver only inpatient hospice or do not distinguish between palliative care and hospice (67%) were conducted in inpatient settings. Sample sizes ranged from 11 to 629 participants (mean 82.9, standard deviation 122.9). This was similar among the hospice model subgroup studies (mean 97.7; standard deviation 136.4).

Fewer than half of all studies (35% total; 32% in the hospice model subgroup) were randomized controlled trials (33-36,40,48,50,51,55,62,65,66). Other study designs included were observational/retrospective cohort designs used to estimate treatment effects or feasibility and uncontrolled clinical trials. Among 19 studies with control groups, thirteen employed usual care or standard palliative care arm (33-36,43,47,50,51,54,55,63,65,67).

Studies’ follow-up periods varied considerably from three days to six months, with an average of 6.8 weeks. Twelve studies did not provide information on the duration of their follow-up period.

Study participants (Table 3) were largely older adults. Of 31 studies that gave participant age information, only three studies had mean or median participant age under 60 years (and only one in the hospice model subgroup) (43,46,67). Most studies recruited more women than men; of 29 studies that gave gender breakdowns, 23 had a higher proportion of female participants. Only ten studies reported the racial and ethnic demographics of their participants. Among those studies providing this information in which white, British white or non-Hispanic white was an option, these were invariably the most highly represented groups in the study.

Interventions and interventionists

The most common types of interventions involved the use of complementary alternative medicine, such as massage, acupuncture, music therapy, and yoga (35% of total studies; 36% of hospice model subgroup studies) rather than conventional BH interventions (psychotherapy or psychopharmacology) (37,38,40,46,48,52,58-63,65). Approximately one-third of the studies involved psychotherapeutic interventions, most frequently cognitive therapy (three studies) (31,35,55) and dignity therapy (two studies) (33,34) (Table 4). Fewer than 20% of interventions involved pharmacologic treatment, most frequently rapid-acting psychotropics such as ketamine (three studies) (44,56,64) and methylphenidate (two studies) (53,66). Of note, all pharmacotherapy studies were conducted in the hospice model subgroup.

While studies often used more than one interventionist, interventions were most often undertaken by research interventionists (16%) (Table 2). Only a minority of studies involved interventions delivered by non-physician mental health clinicians (psychologists, social workers). Other interventionists included nurses (10.8%), physicians (psychiatrists, psycho-oncologists) (10.8%), massage therapists (10.8%), music therapists (8.1%), or interdisciplinary teams (5.4%).

Studies employing session-based interventions ranged in the number of sessions and duration of intervention sessions. Studies ranged in duration from single-session interventions (34,37,40,50-52,56,58) to 28 sessions (in a study of a daily guided self-help intervention) (36). Intervention duration ranged from 20 minutes to 1.5 hours. Ten studies did not provide information about the duration, timing, and frequency of intervention sessions.

Table 3 presents the analyzed studies’ BH outcomes. Studies universally measured symptoms rather than diagnoses (e.g., depressive symptoms rather than major depressive disorder). The most commonly employed measures used to assess treatment effect included the Hospital Anxiety and Depression Scale (HADS) (in 24% of all studies and 32% of hospice model subgroup studies) (81), the Edmonton Symptom Assessment System-Revised (ESAS) (16% total; 14% of hospice model subgroup studies) (82), the Clinical Global Impression Scale (CGI) (8%) (83), and the Patient Health Questionnaire (PHQ) (8%) (84). Depressive symptoms (81% of all studies; 79% of hospice model subgroup studies) and anxiety symptoms (59% of all studies; 57% of hospice model subgroup studies) were the most frequent BH outcomes. Other constructs measured included distress (n=5), mood (n=3), and well-being (n=4). As a note, we distinguished mood from depression scales because the three studies that used mood as an outcome utilized subjective mood scales that asked about a respondent’s overall mood rather than about depression explicitly (e.g., the Memorial Pain Assessment Card asks respondents to rate themselves from “worst mood” to “best mood”) (85). A minority of studies (11%, both hospice model subgroup) employed unvalidated scales or were not specific in their measurement approach. No studies focused on substance use disorders, post-traumatic stress disorder, or serious mental illnesses like psychotic disorders.

Study results

Notable study outcomes are described in Table 3. Slightly fewer than half (17 studies; 46%) of studies demonstrated statistically significant findings related to a BH outcome (i.e., reduction in depression/anxiety symptoms over time between groups) (31,32,34,36,37,40,43,48,54-56,59-63,67). Among the 27 studies in the hospice model subgroup, 9 (33%) had statistically significant results.

Among those studies demonstrating significant results, four significantly improved only anxiety symptoms (36,40,55,59). Two of these studies were complementary and alternative medicine (CAM) interventions, and two were psychotherapy interventions. Two studies significantly improved only depression symptoms (32,37): one CAM study and one psychotherapy intervention. Six studies significantly improved both anxiety and depression symptoms (31,43,61-63,67). Interventions that improved both anxiety and depression symptoms were CAM [3], spiritual care/other [1], combined psychotherapy and CAM [1], and psychotherapy [1]. The remaining five studies improved mood, distress, or overall clinical impression (34,48,54,56,60). These studies consisted of CAM [3], pharmacologic [1], and psychotherapy [1] interventions. Overall, among the 12 studies that significantly improved depression and/or anxiety symptoms, four employed psychotherapy interventions (31,32,36,55), six employed CAM interventions (37,40,59,61-63), one employed combined psychotherapy and complementary and alternative intervention (43), and one employed spiritual care (67).

Among studies that provided a magnitude of change, improvements in depression and anxiety symptoms were relatively modest. For instance, among the two studies utilizing the HADS and demonstrating significant improvements in depression and anxiety symptoms, the magnitude of change on the HADS subscales ranged from 2.7 to 4.7 points (31,61). Of note, these values exceed the proposed minimally important difference of 1.5 points on each component of the HADS (derived from patients with cardiovascular and pulmonary disease) (86,87). Similarly, studies utilizing the ESAS and demonstrating significant improvements in depression and anxiety symptom scores ranged from 1.0 to 2.3 points change on a 10-point scale (59,63). These changes overlap and may exceed the minimally significant difference in individual ESAS symptom measures (88). Most studies measured improvements immediately after the intervention or study, and few demonstrated sustained improvement in symptoms over time following the intervention.

Discussion

We present the findings of a scoping review of BH interventions employed in the hospice setting. Hospice is arguably the most significant model of end-of-life care from a public health perspective. In 2020, over 5,000 hospices provided care to over 1.7 million Americans. Hospice patients experience high rates of BH symptoms, including depression, anxiety, and agitation (9,12). Unmet BH needs negatively impact patients’ quality of life, overall symptom control, and functional status, and diminish caregivers’ quality of life and health (16,89-91).

Despite the prevalence and impact of BH conditions in the hospice setting, we found a striking dearth of research on BH interventions targeting hospice patients. Many of the studies that we did identify were small feasibility studies. Furthermore, many included studies focused only partially on BH outcomes or utilized more general CAM interventions that were not mechanistically focused on BH outcomes (e.g., massage and aromatherapy). We found few randomized controlled trials of psychopharmacologic or psychotherapeutic interventions specifically focused on BH conditions. Furthermore, although most studies looked at depression and/or anxiety symptoms, few focused on delirium/agitation despite its substantial prevalence among hospice patients. Measurement tools were heterogeneous and largely focused on symptom states rather than BH diagnoses. We did not identify any studies on BH interventions for other conditions, such as post-traumatic stress disorder or substance use disorders, despite the burden of these conditions among patients with serious illness (92-94).

The gaps we identified were more pronounced in U.S.-based studies. Among studies conducted in the U.S., we identified only one randomized controlled trial of a psychopharmacologic BH intervention and no study that employed a randomized controlled trial design to test psychotherapy. This is significant because the model of hospice care differs significantly across regions, possibly limiting generalizability. The U.S. model of predominantly home-based care in a capitated payment structure increasingly provided by for-profit private agencies may not translate to other settings (95). For instance, the most represented region in our study was the UK, where a greater proportion of hospice services are provided in inpatient or residential settings (96). Perhaps as a result, included studies underrepresented home-based hospice and over-represented inpatient hospice settings relative to the hospice care landscape in the U.S.

Nearly 40% of included studies involved interventions we categorized as complementary or alternative. Utilization of such interventions may reflect the challenges of deploying traditional BH interventions (psychotherapy and psychopharmacology) in hospice settings, including stigma and prognostic limitations. In addition to being more acceptable to patients and appropriate for the patient population, such interventions may be more feasible in the U.S. hospice care delivery system.

Hospice payment and accountability policies in the U.S. present several barriers to integrating specialist-delivered BH interventions. Hospice care in the U.S. is paid on a capitated basis (that is, hospice organizations are paid per patient per day of care). Payment for the most common form of hospice care, routine home care, hovers between 150–250 U.S. dollars per patient per day, depending on the duration of the patient’s hospice enrollment and the hospice’s adherence to quality reporting guidelines (97). Payments delivered to hospice programs must support a range of requisite services provided by an interdisciplinary team, including a physician, a registered nurse, a social worker, and a spiritual care provider (98). In addition to the inherent challenges that the capitated model presents in providing additional BH services, most hospices operating in the U.S. function in a for-profit model that disincentivizes delivery of non-requisite services (95,99). Hospices participating in quality reporting to the Centers for Medicare & Medicaid Services (CMS) are eligible for higher payment rates. However, hospice quality measures and administrative data do not include BH measures. The consumer survey that constitutes the third quality reporting domain contains 2–4 items about anxiety, sadness, and restlessness/agitation (100). In sum, the care delivery structures in the U.S. hospice system and payment, quality, and accountability policies may disincentivize research and care delivery contingent on BH specialists. In contrast, some complementary and alternative interventions studied, such as aromatherapy or therapeutic music listening, can be delivered using the existing hospice workforce.

Our findings generally emphasize the need for and the challenges of conducting psychopharmacology research among hospice patients. In the U.S., most hospice medication kits contain a small number of medications targeting BH symptoms; these include lorazepam and haloperidol, two of the most commonly prescribed hospice medications (101,102). As many as 80% of hospice patients are prescribed benzodiazepines, and as many as 65% are prescribed antipsychotics (103). Furthermore, many hospice patients may be prescribed antidepressants, though treatment is often interrupted towards the end of life (104). In the absence of high-quality psychopharmacologic trials in the hospice setting, the widespread prescribing and/or deprescribing of psychotropics in the hospice setting is either not evidence-based practice, or, at best, reliant on evidence extrapolated from other settings. The extent to which psychopharmacologic agents are utilized in the hospice setting suggests a critical need for evidence-based guidelines leveraging data from hospice settings.

Our findings further reinforce the discordance between the high utilization of hospice and the challenges in conducting research in hospice settings. Because hospice organizations are often distinct from academic medical institutions and may not have built-in research infrastructure, conducting clinical trials in hospice settings poses unique challenges (4). This is reflected in our data; only one of the seven larger trials (≥100 persons) included in the current study was conducted in the U.S., with the rest conducted in regions with more integration between hospice services and other medical services. However, despite these barriers, building a more robust hospice research infrastructure could provide value to various stakeholders, including payors, patients and caregivers, and clinicians.

Our study contributes to the field of hospice research, focusing specifically on hospice care settings rather than all palliative care more broadly. We identify key research gaps with direct implications for current clinical practice. However, our study has several limitations. The a priori criteria we established in our protocols introduced significant heterogeneity into our study, and thus, our findings must be couched in a broader process-level challenge. Our goal in conducting this study—to explore the landscape of BH research in the hospice setting—was informed by our context as researchers and clinicians in the U.S. Hospice care in the U.S. is a distinct model of care with specific payment and care delivery structures that distinguish it from other forms of palliative and end-of-life care. As such, we elected to limit our search and inclusion criteria to hospice and not include other palliative and end-of-life care models. We specifically chose to privilege the care setting (hospice) over the population (patients at the end of life) because of the distinct challenges of conducting research and implementing interventions in the (U.S.) hospice setting. In conducting our study, we found that much of the terminology that we used did not translate into other health systems. Many health systems interchangeably use language such as palliative care, hospice, and end-of-life care. Furthermore, among systems that distinguish hospice as a specific model of care, the definition of the model may vary significantly and range from a blanket term for palliative care to a model of predominantly in-patient end-of-life care to largely community-based end-of-life care. Many of the nuances, such as the possibility of receiving hospice care concurrently with disease-oriented treatments (which is generally not allowed in U.S. hospice care for adults), differ between settings. As a result, we may have excluded studies that were conducted in applicable settings, perhaps using different terminology, and we included studies that do not reflect the hospice context in which we were interested. We attempted to mitigate this challenge by specifying those studies conducted in the U.S. and those undertaken in systems with similar hospice care structures (the hospice model subgroup); this presents a major limitation and an important process-level finding of our study. In attempting to address this weakness in our study, we were surprised to find the overall shortage of descriptions of hospice as a care model across settings. In this sense, a key recommendation emerging from our study is the need for literature describing models of hospice care across health settings to ensure that researchers conducting studies such as ours oriented around a health delivery model can appropriately include studies conducted globally.

In addition to the systems-level heterogeneity, our study has several other limitations. Our objectives of understanding the research landscape led us to be broad in including studies. As such, some included studies were nebulous regarding whether their design met inclusion criteria and whether they were sufficiently focused on BH outcomes. We also decided not to appraise study quality, which is a limitation in using our work to generate clinical recommendations. However, we felt that study appraisal was inconsistent with our objective of broadly describing existing literature. Finally, as is often the case in hospice and palliative care research, our studies overrepresented cancer patients; 70% of our studies included predominantly or exclusively patients with cancer diagnoses.

Our study represents a clarion call for greater investment in hospice research infrastructure and the use of clinical trials to address the pressing needs of hospice patients. We identified a striking lack of BH intervention research focused on hospice patients compared to specialized serious illness care like oncology or other palliative care models. Addressing this gap requires a multifactorial approach. Investing in research infrastructure and partnerships between private hospices and academic medical institutions could help facilitate research in hospice settings. This is of particular importance given the unique features of the U.S. hospice care model. Further investment in general BH-palliative care integration is also critical. In particular, there is a need to increase the BH workforce and leverage existing BH experts such as hospice and palliative care social workers as investigators and clinicians supported in carrying out BH assessments and interventions (10,105).

Conclusions

We conducted a scoping review of studies focused on BH interventions in the hospice setting. We identified 37 studies from hospice settings in 16 regions, including complementary and alternative, psychotherapeutic, and psychopharmacologic interventions. Most studies were small pilot and feasibility studies, while few studies employed adequately powered clinical trials to test for treatment effects. In addition, relatively few studies took place in the U.S., and these studies tended to be smaller, possibly reflecting structural barriers to conducting hospice research in the U.S. Our study highlights two significant gaps in hospice research. First, we identified substantial heterogeneity in terminology around hospice services between health systems, making it difficult to make inferences about research conducted between different health systems. This finding highlights a need for resources to help scholars and health service users across settings understand similarities and differences across models of care in different health systems. Many existing studies we identified were conducted in settings where hospice care was analogous to palliative care more generally, rather than settings with a distinct hospice model of care different from general palliative care services. Second, our study identifies a significant gap in hospice research. Despite the public health importance of the hospice model and the prevalence of impactful BH conditions in patients receiving hospice care, there are few randomized controlled trials of BH interventions for hospice patients. As such, current BH practices in hospice settings, including high utilization of benzodiazepines and antipsychotics, may not be driven by evidence or represent optimal end-of-life BH care.

Acknowledgments

Funding: This study was supported by the Weill Cornell JumpStart Research Award Program (to D.S.), and the National Institute on Aging award numbers K24-AG053462 (to M.C.R.) and T32AG049666 (to D.S., M.C.R.).

Footnote

Reporting Checklist: The authors have completed the PRISMA-ScR reporting checklist. Available at https://apm.amegroups.com/article/view/10.21037/apm-23-508/rc

Peer Review File: Available at https://apm.amegroups.com/article/view/10.21037/apm-23-508/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://apm.amegroups.com/article/view/10.21037/apm-23-508/coif). M.C.R. reports receiving funding from National Institute on Aging for this study. D.S. reports receiving funding from National Institute on Aging and Weill Cornell JumpStart Research program for this study. The other authors have no conflicts of interest to disclose.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Ferrell BR, Twaddle ML, Melnick A, et al. National Consensus Project Clinical Practice Guidelines for Quality Palliative Care Guidelines, 4th Edition. J Palliat Med 2018;21:1684-9.
2. Sheikh M, Sekaran S, Kochhar H, et al. Hospice vs Palliative care: A comprehensive review for primary care physician. J Family Med Prim Care 2022;11:4168-73. [Crossref] [PubMed]
3. Hospice Facts & Figures [Internet]. NHPCO. [cited 2021 Feb 16]. Available online: https://www.nhpco.org/factsfigures/
4. Oliver DP, Washington KT, Demiris G, et al. Challenges in Implementing Hospice Clinical Trials: Preserving Scientific Integrity While Facing Change. J Pain Symptom Manage 2020;59:365-71. [Crossref] [PubMed]
5. American Medical Association [Internet]. 2022 [cited 2023 Jul 31]. What is behavioral health? Available online: https://www.ama-assn.org/delivering-care/public-health/what-behavioral-health
6. Austin P, Wiley S, McEvoy PM, et al. Depression and anxiety in palliative care inpatients compared with those receiving palliative care at home. Palliat Support Care 2011;9:393-400. [Crossref] [PubMed]
7. Sewtz C, Muscheites W, Grosse-Thie C, et al. Longitudinal observation of anxiety and depression among palliative care cancer patients. Ann Palliat Med 2021;10:3836-46. [Crossref] [PubMed]
8. van Oorschot B, Ishii K, Kusomoto Y, et al. Anxiety, depression and psychosocial needs are the most frequent concerns reported by patients: preliminary results of a comparative explorative analysis of two hospital-based palliative care teams in Germany and Japan. J Neural Transm (Vienna) 2020;127:1481-9. [Crossref] [PubMed]
9. Kozlov E, Phongtankuel V, Prigerson H, et al. Prevalence, Severity, and Correlates of Symptoms of Anxiety and Depression at the Very End of Life. J Pain Symptom Manage 2019;58:80-5. [Crossref] [PubMed]
10. Cheung S, Spaeth-Rublee B, Shalev D, et al. A Model to Improve Behavioral Health Integration into Serious Illness Care. J Pain Symptom Manage 2019;58:503-514.e1. [Crossref] [PubMed]
11. Kozlov E, Dong X, Kelley AS, et al. The Epidemiology of Depressive Symptoms in the Last Year of Life. J Am Geriatr Soc 2020;68:321-8. [Crossref] [PubMed]
12. Recchia A, Rizzi B, Favero A, et al. Prevalence of Delirium in End-of-Life Palliative Care Patients: An Observational Study. Med Princ Pract 2022;31:118-24. [Crossref] [PubMed]
13. Kozlov E, Eghan C, Moran S, et al. Palliative Care Providers' Practices Surrounding Psychological Distress Screening and Treatment: A National Survey. Am J Hosp Palliat Care 2018;35:938-44. [Crossref] [PubMed]
14. Kozlov E, Niknejad B, Reid MC. Palliative Care Gaps in Providing Psychological Treatment: A Review of the Current State of Research in Multidisciplinary Palliative Care. Am J Hosp Palliat Care 2018;35:505-10. [Crossref] [PubMed]
15. Nowels MA, Kalra S, Duberstein PR, et al. Palliative Care Interventions Effects on Psychological Distress: A Systematic Review & Meta-Analysis. J Pain Symptom Manage 2023;65:e691-e713. [Crossref] [PubMed]
16. Sadowska K, Fong T, Horning DR, et al. Psychiatric Comorbidities and Outcomes in Palliative and End-of-Life Care: A Systematic Review. J Pain Symptom Manage 2023;66:e129-51. [Crossref] [PubMed]
17. Atkin N, Vickerstaff V, Candy B. 'Worried to death': the assessment and management of anxiety in patients with advanced life-limiting disease, a national survey of palliative medicine physicians. BMC Palliat Care 2017;16:69. [Crossref] [PubMed]
18. Brenne E, Loge JH, Lie H, et al. The Edmonton Symptom Assessment System: Poor performance as screener for major depression in patients with incurable cancer. Palliat Med 2016;30:587-98. [Crossref] [PubMed]
19. Shalev D, Patterson M, Aytaman Y, et al. Operationalizing Depression Screening in Ambulatory Palliative Care: A Quality Improvement Project. J Pain Symptom Manage 2023;65:e7-e13. [Crossref] [PubMed]
20. Rayner L, Price A, Hotopf M, et al. Expert opinion on detecting and treating depression in palliative care: A Delphi study. BMC Palliat Care 2011;10:10. [Crossref] [PubMed]
21. Perusinghe M, Chen KY, McDermott B. Evidence-Based Management of Depression in Palliative Care: A Systematic Review. J Palliat Med 2021;24:767-81. [Crossref] [PubMed]
22. Palliative Care Beyond End-of-Life Care | The Joint Commission [Internet]. [cited 2023 Aug 11]. Available online: https://www.jointcommission.org/resources/news-and-multimedia/blogs/improvement-insights/2022/04/palliative-care-beyond-end-of-life-care/
23. McInnerney D, Candy B, Stone P, et al. Access to and adequacy of psychological services for adult patients in UK hospices: a national, cross-sectional survey. BMC Palliat Care 2021;20:31. [Crossref] [PubMed]
24. Allsop MJ, Ziegler LE, Mulvey MR, et al. Duration and determinants of hospice-based specialist palliative care: A national retrospective cohort study. Palliat Med 2018;32:1322-33. [Crossref] [PubMed]
25. Jordan RI, Allsop MJ, ElMokhallalati Y, et al. Duration of palliative care before death in international routine practice: a systematic review and meta-analysis. BMC Med 2020;18:368. [Crossref] [PubMed]
26. Tricco AC, Lillie E, Zarin W, et al. PRISMA Extension for Scoping Reviews (PRISMA-ScR): Checklist and Explanation. Ann Intern Med 2018;169:467-73. [Crossref] [PubMed]
27. Arksey H, O’Malley L. Scoping studies: towards a methodological framework. International Journal of Social Research Methodology 2005;8:19-32. [Crossref]
28. Westphaln KK, Regoeczi W, Masotya M, et al. From Arksey and O’Malley and Beyond: Customizations to enhance a team-based, mixed approach to scoping review methodology. MethodsX 2021;8:101375. [Crossref] [PubMed]
29. Complementary and Alternative Medicine. In: Collection Development Guidelines of the National Library of Medicine [Internet] [Internet]. National Library of Medicine (US); 2018 [cited 2023 Aug 13]. Available online: https://www.ncbi.nlm.nih.gov/books/NBK518811/
30. Covidence [Internet]. [cited 2022 Dec 5]. Covidence - Better systematic review management. Available online: https://www.covidence.org/
31. Anderson T, Watson M, Davidson R. The use of cognitive behavioural therapy techniques for anxiety and depression in hospice patients: a feasibility study. Palliat Med 2008;22:814-21. [Crossref] [PubMed]
32. Cheng JO, Lo R, Chan F, et al. A pilot study on the effectiveness of anticipatory grief therapy for elderly facing the end of life. J Palliat Care 2010;26:261-9. [Crossref] [PubMed]
33. Chochinov HM, Kristjanson LJ, Breitbart W, et al. Effect of dignity therapy on distress and end-of-life experience in terminally ill patients: a randomised controlled trial. Lancet Oncol 2011;12:753-62. [Crossref] [PubMed]
34. De Vincenzo F, Lombardo L, Iani L, et al. Spiritual well-being, dignity-related distress and demoralisation at the end of life-effects of dignity therapy: a randomised controlled trial. BMJ Support Palliat Care 2023;bmjspcare-2022-003696.
35. Fang CK, Li PY. Existential cognitive therapy for terminal cancer patients with depression or demoralization: A randomized controlled trial. Psycho-Oncology 2017;26:60-1.
36. Galfin JM, Watkins ER, Harlow T. A brief guided self-help intervention for psychological distress in palliative care patients: a randomised controlled trial. Palliat Med 2012;26:197-205. [Crossref] [PubMed]
37. Gallagher LM, Lagman R, Bates D, et al. Perceptions of family members of palliative medicine and hospice patients who experienced music therapy. Support Care Cancer 2017;25:1769-78. [Crossref] [PubMed]
38. Havyer RD, Lapid MI, Dockter TJ, et al. Impact of Massage Therapy on the Quality of Life of Hospice Patients and Their Caregivers: A Pilot Study. J Palliat Care 2022;37:41-7. [Crossref] [PubMed]
39. Henderson M, MacGregor E, Sykes N, et al. The use of benzodiazepines in palliative care. Palliat Med 2006;20:407-12. [Crossref] [PubMed]
40. Horne-Thompson A, Grocke D. The effect of music therapy on anxiety in patients who are terminally ill. J Palliat Med 2008;11:582-90. [Crossref] [PubMed]
41. Hudson SF, Campbell F, Leek J, et al. Space to breathe: A new hospice based palliative care, respiratory and psychology programme for patients with severe COPD and their carers. Thorax 2013;68:A204. [Crossref]
42. Hulbert-Williams NJ, Norwood SF, Gillanders D, et al. Brief Engagement and Acceptance Coaching for Hospice Settings (the BEACHeS study): results from a Phase I study of acceptability and initial effectiveness in people with non-curative cancer. BMC Palliat Care 2021;20:96. [Crossref] [PubMed]
43. Husić S, Mesić D. Daily hospice: Depression and anxiety after mastectomy for breast cancer. Acta Clin Croat 2010;49:19-24. [PubMed]
44. Iglewicz A, Morrison K, Nelesen RA, et al. Ketamine for the treatment of depression in patients receiving hospice care: a retrospective medical record review of thirty-one cases. Psychosomatics 2015;56:329-37. [Crossref] [PubMed]
45. Imrie S, Troop NA. A pilot study on the effects and feasibility of compassion-focused expressive writing in Day Hospice patients. Palliat Support Care 2012;10:115-22. [Crossref] [PubMed]
46. Johnston B, Bowman F, Carduff E, et al. “Playlist for Life” at the end of life: a mixed-methods feasibility study of a personalised music listening intervention in the hospice setting. Pilot and Feasibility Studies 2022;8:32. [Crossref] [PubMed]
47. Kieszkowska-Grudny A, Jarosz J, Kaczmarek Z, et al. The place and role of Skype consultancies among palliative patients and the impact of this type of care on a quality of life, pain, anxiety and depression symptoms assessment in home hospice care patient. J Clin Oncol 2016;34:e21512. [Crossref]
48. Kutner JS, Smith MC, Corbin L, et al. Massage therapy versus simple touch to improve pain and mood in patients with advanced cancer: a randomized trial. Ann Intern Med 2008;149:369-79. [Crossref] [PubMed]
49. Lloyd-Williams M, Payne S, Reeve J, et al. Antidepressant medication in patients with advanced cancer--an observational study. QJM 2013;106:995-1001. [Crossref] [PubMed]
50. Lloyd-Williams M, Cobb M, O'Connor C, et al. A pilot randomised controlled trial to reduce suffering and emotional distress in patients with advanced cancer. J Affect Disord 2013;148:141-5. [Crossref] [PubMed]
51. Lloyd-Williams M, Shiels C, Ellis J, et al. Pilot randomised controlled trial of focused narrative intervention for moderate to severe depression in palliative care patients: DISCERN trial. Palliat Med 2018;32:206-15. [Crossref] [PubMed]
52. Louis M, Kowalski SD. Use of aromatherapy with hospice patients to decrease pain, anxiety, and depression and to promote an increased sense of well-being. Am J Hosp Palliat Care 2002;19:381-6. [Crossref] [PubMed]
53. Macleod AD. Methylphenidate in terminal depression. J Pain Symptom Manage 1998;16:193-8. [PubMed]
54. Moon H, Journ S, Lee S. Effect of Laughter Therapy on Mood Disturbances, Pain, and Burnout in Terminally Ill Cancer Patients and Family Caregivers. Cancer Nurs 2024;47:3-11. [Crossref] [PubMed]
55. Moorey S, Cort E, Kapari M, et al. A cluster randomized controlled trial of cognitive behaviour therapy for common mental disorders in patients with advanced cancer. Psychol Med 2009;39:713-23. [Crossref] [PubMed]
56. Morrison K, Irwin S, Nelesen R, et al. A retrospective chart review of the efficacy of ketamine for depression in patients receiving hospice care (FR416-A). Journal of Pain and Symptom Management 2013;45:373. [Crossref]
57. Myrcik D, Statowski W, Trzepizur M, et al. Influence of Physical Activity on Pain, Depression and Quality of Life of Patients in Palliative Care: A Proof-of-Concept Study. J Clin Med 2021;10:1012. [Crossref] [PubMed]
58. Nakayama H, Kikuta F, Takeda H. A pilot study on effectiveness of music therapy in hospice in Japan. J Music Ther 2009;46:160-72. [Crossref] [PubMed]
59. Pedersen K, Björkhem-Bergman L. Tactile massage reduces rescue doses for pain and anxiety: an observational study. BMJ Support Palliat Care 2018;8:30-3. [Crossref] [PubMed]
60. Pilgrem E. Social and therapeutic horticulture as a palliative care intervention. BMJ Support Palliat Care 2023;13:323-6. [Crossref] [PubMed]
61. Plaskota M, Lucas C, Evans R, et al. A hypnotherapy intervention for the treatment of anxiety in patients with cancer receiving palliative care. Int J Palliat Nurs 2012;18:69-75. [Crossref] [PubMed]
62. Ramanathan M, Bhavanani AB, Trakroo M. Effect of a 12-week yoga therapy program on mental health status in elderly women inmates of a hospice. Int J Yoga 2017;10:24-8. [Crossref] [PubMed]
63. Romeo MJ, Parton B, Russo RA, et al. Acupuncture to Treat the Symptoms of Patients in a Palliative Care Setting. Explore (NY) 2015;11:357-62. [Crossref] [PubMed]
64. Shea K, Pucko J, Koon M, et al. Improving end-of-life quality through the implementation and testing of a ketamine protocol for reduction of depression and pain. Journal of Pain and Symptom Management 2018;55:561-2. [Crossref]
65. Soden K, Vincent K, Craske S, et al. A randomized controlled trial of aromatherapy massage in a hospice setting. Palliat Med 2004;18:87-92. [Crossref] [PubMed]
66. Sullivan DR, Mongoue-Tchokote S, Mori M, et al. Randomized, double-blind, placebo-controlled study of methylphenidate for the treatment of depression in SSRI-treated cancer patients receiving palliative care. Psychooncology 2017;26:1763-9. [Crossref] [PubMed]
67. Sun XH, Liu X, Zhang B, et al. Impact of spiritual care on the spiritual and mental health and quality of life of patients with advanced cancer. World J Psychiatry 2021;11:449-62. [Crossref] [PubMed]
68. Hospice UK [Internet]. [cited 2023 Oct 31]. What is hospice care? Available online: https://www.hospiceuk.org/information-and-support/your-guide-hospice-and-end-life-care/im-looking-hospice-care/what-hospice
69. Bradley NM, Dowrick CF, Lloyd-Williams M. A survey of hospice day services in the United Kingdom & Republic of Ireland : how did hospices offer social support to palliative care patients, pre-pandemic? BMC Palliat Care 2022;21:170. [Crossref] [PubMed]
70. Rees-Roberts M, Williams P, Hashem F, et al. Hospice at Home services in England: a national survey. BMJ Support Palliat Care 2021;11:454-60. [Crossref] [PubMed]
71. Department of Health & Human Services. End of life and palliative care explained [Internet]. Department of Health & Human Services; [cited 2023 Oct 31]. Available online: http://www.betterhealth.vic.gov.au/health/servicesandsupport/end-of-life-and-palliative-care-explained
72. Hospice Care – Hospice New Zealand [Internet]. [cited 2023 Oct 31]. Available online: https://www.hospice.org.nz/what-is-hospice/
73. Hospice | CMS [Internet]. [cited 2023 Sep 2]. Available online: https://www.cms.gov/medicare/medicare-fee-for-service-payment/hospice
74. Canadian Hospice Palliative Care Association [Internet]. [cited 2023 Oct 31]. Hospice Palliative Care. Available online: https://www.chpca.ca/about-hpc/
75. Aniśko-Trembecka P, Popławska M, Krajewska-Kułak E, et al. The Functioning of Hospice in the Perception of Family Members of Cancer Surgery and Hospice Patients. Int J Environ Res Public Health 2023;20:5334. [Crossref] [PubMed]
76. Lu Y, Gu Y, Yu W. Hospice and Palliative Care in China: Development and Challenges. Asia Pac J Oncol Nurs 2018;5:26-32. [Crossref] [PubMed]
77. Mori M, Morita T. Advances in Hospice and Palliative Care in Japan: A Review Paper. The Korean Journal of Hospice and Palliative Care 2016;19:283-91. [Crossref]
78. Lee CY, Komatsu H, Zhang W, et al. Comparison of the hospice systems in the United States, Japan and taiwan. Asian Nurs Res (Korean Soc Nurs Sci) 2010;4:163-73. [Crossref] [PubMed]
79. Shukla R, Singh N, Acharya S, et al. Hospice care in India: A review. J Family Med Prim Care 2022;11:4987-90. [Crossref] [PubMed]
80. Aebischer Perone S, Nikolic R, Lazic R, et al. Addressing the needs of terminally-ill patients in Bosnia-Herzegovina: patients’ perceptions and expectations. BMC Palliat Care 2018;17:123. [Crossref] [PubMed]
81. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand 1983;67:361-70. [Crossref] [PubMed]
82. Chang VT, Hwang SS, Feuerman M. Validation of the Edmonton Symptom Assessment Scale. Cancer 2000;88:2164-71. [Crossref] [PubMed]
83. Busner J, Targum SD. The Clinical Global Impressions Scale. Psychiatry (Edgmont) 2007;4:28-37. [PubMed]
84. Kroenke K, Spitzer RL, Williams JBW. The PHQ-9. J Gen Intern Med 2001;16:606-13. [Crossref] [PubMed]
85. Fishman B, Pasternak S, Wallenstein SL, et al. The Memorial Pain Assessment Card. A valid instrument for the evaluation of cancer pain. Cancer 1987;60:1151-8. [Crossref] [PubMed]
86. Puhan MA, Frey M, Büchi S, et al. The minimal important difference of the hospital anxiety and depression scale in patients with chronic obstructive pulmonary disease. Health Qual Life Outcomes 2008;6:46. [Crossref] [PubMed]
87. Lemay KR, Tulloch HE, Pipe AL, et al. Establishing the Minimal Clinically Important Difference for the Hospital Anxiety and Depression Scale in Patients With Cardiovascular Disease. J Cardiopulm Rehabil Prev 2019;39:E6-E11. [Crossref] [PubMed]
88. Hui D, Shamieh O, Paiva CE, et al. Minimal clinically important differences in the Edmonton Symptom Assessment Scale in cancer patients: A prospective, multicenter study. Cancer 2015;121:3027-35. [Crossref] [PubMed]
89. Segrin C, Badger TA, Harrington J. Interdependent psychological quality of life in dyads adjusting to prostate cancer. Health Psychol 2012;31:70-9. [Crossref] [PubMed]
90. Schulz R, Beach SR, Hebert RS, et al. Spousal suffering and partner's depression and cardiovascular disease: the Cardiovascular Health Study. Am J Geriatr Psychiatry 2009;17:246-54. [Crossref] [PubMed]
91. Monin JK, Schulz R. Interpersonal effects of suffering in older adult caregiving relationships. Psychol Aging 2009;24:681-95. [Crossref] [PubMed]
92. Bickel KE, Kennedy R, Levy C, et al. The Relationship of Post-traumatic Stress Disorder to End-of-life Care Received by Dying Veterans: a Secondary Data Analysis. J Gen Intern Med 2020;35:505-13. [Crossref] [PubMed]
93. Chhatre S, Metzger DS, Malkowicz SB, et al. Substance use disorder and its effects on outcomes in men with advanced-stage prostate cancer. Cancer 2014;120:3338-45. [Crossref] [PubMed]
94. Ho P, Rosenheck R. Substance Use Disorder Among Current Cancer Patients: Rates and Correlates Nationally in the Department of Veterans Affairs. Psychosomatics 2018;59:267-76. [Crossref] [PubMed]
95. Aldridge MD, Schlesinger M, Barry CL, et al. National hospice survey results: for-profit status, community engagement, and service. JAMA Intern Med 2014;174:500-6. [Crossref] [PubMed]
96. Remington R, Wakim G. A comparison of hospice in the United States and the United Kingdom: implications for policy and practice. J Gerontol Nurs 2010;36:16-21. [Crossref] [PubMed]
97. CMS-1787-F | CMS [Internet]. [cited 2023 Oct 31]. Available online: https://www.cms.gov/medicare/medicare-fee-service-payment/hospice/hospice-regulations-and-notices/1041017703/cms-1787-f
98. 42 CFR Part 418 -- Hospice Care [Internet]. [cited 2023 Oct 31]. Available online: https://www.ecfr.gov/current/title-42/part-418
99. Odejide OO, Aldridge MD. Access to High-Quality Hospice Care in a For-Profit World. Oncologist 2023;28:743-5. [Crossref] [PubMed]
100. CAHPS® Hospice Survey | CMS [Internet]. [cited 2023 Oct 31]. Available online: https://www.cms.gov/medicare/quality/hospice/cahpsr-hospice-survey
101. Bishop MF, Stephens L, Goodrich M, et al. Medication kits for managing symptomatic emergencies in the home: a survey of common hospice practice. J Palliat Med 2009;12:37-44. [Crossref] [PubMed]
102. Gerlach LB, Powell V, Zhang L, et al. The Most Common Medications Billed to the Medicare Hospice Benefit Among Hospice Beneficiaries in the U.S. J Pain Symptom Manage 2021;62:e3-6. [Crossref] [PubMed]
103. Gerlach LB, Zhang L, Strominger J, et al. Hospice agency characteristics associated with benzodiazepine and antipsychotic prescribing. J Am Geriatr Soc 2023;71:2571-8. [Crossref] [PubMed]
104. Crampton E, Weisse CS. Antidepressant Use During Hospice Patients' Final Months on Routine Home Care. Am J Hosp Palliat Care 2023;40:953-8. [Crossref] [PubMed]
105. Shalev D, Docherty M, Spaeth-Rublee B, et al. Bridging the Behavioral Health Gap in Serious Illness Care: Challenges and Strategies for Workforce Development. Am J Geriatr Psychiatry 2020;28:448-62. [Crossref] [PubMed]