Care transitions and social needs: A Geriatric Emergency care Applied Research (GEAR) Network scoping review and consensus statement

Abstract

Objectives:

Individual-level social needs have been shown to substantially impact emergency department (ED) care transitions of older adults. The Geriatric Emergency care Applied Research (GEAR) Network aimed to identify care transition interventions, particularly addressing social needs, and prioritize future research questions.

Methods:

GEAR engaged 49 interdisciplinary stakeholders, derived clinical questions, and conducted searches of electronic databases to identify ED discharge care transition interventions in older adult populations. Informed by the Protocol for Responding to and Assessing Patients’ Assets, Risks, and Experiences (PRAPARE) framework, data extraction and synthesis of included studies included the degree that intervention components addressed social needs and their association with patient outcomes. GEAR convened a consensus conference to identify topics of highest priority for future care transitions research.

Results:

Our search identified 248 unique articles addressing care transition interventions in older adult populations. Of these, 17 individual care transition intervention studies were included in the current literature synthesis. Overall, common care transition interventions included coordination efforts, comprehensive geriatric assessments, discharge planning, and telephone or in-person follow-up. Fourteen of the 17 care transition intervention studies in older adults specifically addressed at least one social need within the PRAPARE framework, most commonly related to access to food, medicine, or health care. No care transition intervention addressing social needs in older adult populations consistently reduced subsequent health care utilization or other patient-centered outcomes. GEAR stakeholders identified that determining optimal outcome measures for ED–home transition interventions was the highest priority area for future care transitions research.

Conclusions:

ED care transition intervention studies in older adults frequently address at least one social need component and exhibit variation in the degree of success on a wide array of health care utilization outcomes.

INTRODUCTION

Older adults, defined as those aged 65 years and older, account for over 22 million emergency department (ED) visits annually, representing 16% of all ED visits nationally.¹ Approximately 65% of older adult ED patients are discharged home.² This vulnerable care transition is often fraught with inadequate communication and poor patient comprehension of their medical condition.^3–6 Deficiencies in ED care transitions have been associated with an increased likelihood of adverse health outcomes including functional status decline, unscheduled ED revisitation, hospital admission, and mortality.^7–11 Moreover, there is increasing awareness that social determinants of health (SDOH) have a significant effect on care transitions after the receipt of acute unscheduled care.^12–14 Defined by the World Health Organization,¹⁵ community-level SDOH are “the non-medical factors that influence health outcomes,” frequently encompassing education, employment, food security, housing, social inclusion and nondiscrimination, and access to affordable health services. Social risk factors are individual-level adverse SDOH that are often identified through screening tools (e.g., positive for food insecurity), and social needs are the downstream individual-level needs prioritized by patients (e.g., request for food assistance).^16–18 While social risk factors have been noted to contribute to ED use and adverse outcomes,^19–21 EDs also face significant challenges in identifying and intervening on social needs.^22,23

Particularly for older adults, the intersection of ED care transitions and SDOH represents a key quality focus. Their contemporary relevance is evidenced by priority within the 2014 Geriatric ED Guidelines,²⁴ the 2017 National Quality Forum ED Transitions of Care Report,²⁵ the 2018 launch of the American College of Emergency Physicians tiered geriatric ED accreditation process,²⁶ the growing role of social emergency medicine,^27–30 and the Social Interventions Research & Evaluation Network (SIREN).³¹ Despite this increased interest and a multitude of observational studies that have documented high rates of social needs in older ED patients,^32–34 it is unclear whether care transition interventions address social needs in older adults and, if so, whether these interventions are effective. Prior reviews of care transition interventions have been limited by focusing on a general population of adult patients (e.g., including, but not limited to, geriatric patients) as well as a lack of accounting for a broad and inclusive range of social needs that are increasingly recognized to play a significant role in successful care transitions.^22,35–37 Knowledge of prior successes and identification of key research questions going forward will be crucial to implement effective patient-centered interventions in the growing older adult population.

In 2019, the Geriatric Emergency care Applied Research (GEAR) Network convened an interdisciplinary group of stakeholders to identify and address high-impact research questions in multiple areas, including care transitions. Collectively, this group of clinicians and researchers identified the need to consider the impact of social needs on care transitions following ED discharge, particularly for older adults. Therefore, we conducted a scoping literature review using a patient–intervention–control–outcome (PICO) framework to identify the role that social needs play in previously studied ED discharge care transition interventions targeting older adults. Specifically, we aimed to describe the type, frequency, and associated clinical outcomes of ED care transition interventions that have addressed social needs in older adults. We also aimed to identify high-yield questions for future geriatric emergency care research focused on care transitions.

METHODS

Study design

Participating GEAR Network members were identified by membership in geriatric and emergency medicine interest groups (Society for Academic Emergency Medicine, American Geriatrics Society, and the Gerontological Society of America) as well as their prior pertinent publication record. The GEAR Care Transitions (GEAR CT) subgroup included one geriatrician (SH), six emergency physicians (CG, LDR, TH, JSu, LR, CB), one social psychologist (CV), two nurse scientists (AABri, AABro), one public health professional (MG), one research scientist (KJK), one research assistant (NH), and one social welfare scientist (CM). To enhance the methodological quality, we based the review on the stepwise framework suggested by Levac et al.³⁸

Building consensus and search strategy

The GEAR CT subgroup performed a scoping review that adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews (PRISMA-ScR) reporting guidelines.³⁹ No protocol has been published. As a first step, the 12 GEAR CT members participated in monthly teleconference calls to identify considered PICO questions of interest. Together with a medical librarian, GEAR CT subgroup members developed a literature search strategy and identified exemplar articles. These were presented to 49 clinicians, investigators, and other stakeholders including patient and caregiver representatives participating at an in-person GEAR Consensus Conference in October 2019. Based on discussion regarding the importance of SDOH on ED care transitions with the broader GEAR Network, the GEAR CT subgroup proceeded with a review using the following PICO framework:

Population: ≥65-year-old ED patients discharged to another setting of care or home;

Intervention: Discharge planning/transition/care coordination, transfer forms/Web portals, telephone or in-person follow-up after ED visit that includes referral to community services for health and social needs;

Comparison: Usual care, historical controls/preintervention group;

Outcome: Health care utilization (repeat ED visits, hospital admissions, nursing home entry, follow-up appointments), mortality, costs, patient-reported outcomes (e.g., function, quality of life, patient/caregiver experience, provider experience).

Electronic searches were conducted by a medical librarian within OVID MEDLINE, Embase, CINAHL, and Cochrane Central Register of Controlled Trials from inception through April 2021. Full details of the search terms are provided in Data Supplement S1, Appendix S1 (available as supporting information in the online version of this paper, which is available at http://onlinelibrary.wiley.com/doi/10.1111/acem.14360/full).

Study selection

Studies were identified using a two-phase process. In phase I, reviewers first assessed systematic reviews given the extensive and complex body-of-care transitions research, allowing the GEAR CT subgroup to address a broad research question. Two reviewers screened titles and abstracts identified through our initial search that met the following eligibility criteria: (1) use a systematic review design, (2) include ED-based studies, and (3) evaluate studies of interventions as defined earlier. Citations identified by either of the two reviewers underwent full-text screening. In phase II, reviewers screened titles and abstracts of primary publications that were included in one or more of the identified systematic reviews to determine if they met the following eligibility criteria: (1) include older adults (≥65 years of age) presenting to the ED and discharged to another setting of care or home, (2) evaluate studies of interventions as defined earlier, (3) report on a comparator reference standard, and (4) include at least one outcome of interest. Outcomes of interest included health care utilization (e.g., repeat ED visits, hospital admissions, nursing home entry), costs, and patient-reported outcomes (e.g., function; quality of life; patient, caregiver, or provider experience). Citations identified by either reviewer underwent full-text screening, and discrepancies were resolved by group discussion. A preplanned adjudicator (UH) resolved continued inter-rater discrepancies.

Data abstraction

GEAR CT subgroup members abstracted key study data from filtered results and synthesized findings. Pairs of authors independently reviewed the full text of eligible phase II studies. Data abstracted included the study setting, inclusion/exclusion criteria, study design, comparator reference standard, and primary/secondary outcomes. Similar to prior research methodology,^40–41 a single reviewer independently abstracted the key data elements with confirmation by a second reviewer. To ascertain which, if any, social needs were addressed in these studies, we used the Protocol for Responding to and Assessing Patients’ Assets, Risks, and Experiences (PRAPARE) tool designed to identify and act on SDOHs.^42,43 These included the presence of an intervention component focused on one or more of the following: (1) housing, money, work, or insurance; (2) access to food, medicine, health care, or childcare; (3) transportation; (4) safety, violence, partner, or elder abuse; (5) social support or loneliness; (6) stress or emotional health; (7) health behaviors; or (8) health information, education, or literacy. For any intervention that addressed a listed social need, we also recorded details on whether the intervention included an assessment or referral and, if so, to what type of agency or service. We classified interventions according to social need components identified (none vs. any) and assessed outcome patterns.

In accordance with guidelines for scoping reviews, we aimed to provide an overview of the existing evidence and review different types of evidence.⁴⁴ We further followed the suggestion of Silverman⁴⁵ and synthesized evidence with the support of a table to enhance reliability. Due to the scoping review approach, as well as the heterogeneity of included articles, a formal assessment of study quality or risk of bias was not completed.

RESULTS

Evidence synthesis

After duplicates were removed, phase I of the literature search identified 354 potential systematic reviews, 13 of which met phase I eligibility criteria. Phase II of the literature identified 248 unique articles within any of the 13 systematic reviews, with 25 original research articles identified for full-text review. After full-text review, adjudication, and abstraction, 18 papers describing 17 individual studies met phase II eligibility criteria and were included in the scoping review (Figure 1).

FIGURE 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews flow diagram showing the different steps of study selection

Study characteristics

Characteristics of the 13 systematic reviews are shown in Table s1.^40,46–57 Of the 17 individual studies, all were published between 1996 and 2018. Among the 17 studies, eight were randomized controlled trials (RCTs),^58–65 three were pre/post studies,^66–68 two were case–control studies,^69,70 and two were nonrandomized clinical trials.^71,72 There was also one quasi-RCT⁷³ and one prospective study with in-group comparison.⁷⁴ The sum of study populations totaled 20,837 individuals, with all participants ≥65 years of age (Table S2).

Intervention characteristics

Building from the PICO framework, intervention types were grouped into five different categories, with many studies incorporating interventions with multiple components, thereby identified within more than one category. The first category consisted of eight studies incorporating care coordination efforts and included interventions identifying and addressing barriers to follow-up with needed services.^{61,62,64,66,69–72} The second category consisted of brief or comprehensive geriatric assessments (CGA), included within seven studies.^{60,62,63,65,70,71,74} The third category consisted of four intervention studies targeting discharge planning, with efforts primarily assessing patient comprehension and providing referral to needed services.^63,67–69 The fourth and fifth categories, respectively, included five studies each that identified in-person follow-up^{61,64,71,73,74} or telephone follow-up^{58,59,61,63,71} as part of the intervention to improve ED care transitions for older adults.

Using the framework of the PRAPARE tool, 14 of the 17 studies using care transition interventions addressed at least one social need (Table S3).^{58,59,62–66,68–74} Of these, three studies addressed housing, money, work, or insurance,^63,68,71 while all 14 studies addressed some component of access to food, medicine, health care, or childcare. Within the studies addressing social needs, interventions providing referrals to home services (n = 7),^{58,63,65,68,70,71,73} ensuring access to care through follow-up appointments (n = 5),^{58,59,62,65,73} and providing medication access and reconciliation (n = 4)^58,59,72,74 were the most common. One study addressed transportation needs⁶⁸; seven studies addressed social support or loneliness^{64,66,69,71,74}; and two studies included an assessment of safety, violence, partner, or elder abuse within the intervention.^63,69 Finally, four interventions addressed stress or emotional health^63,68,71,74; six studies addressed health behaviors (including exercise, tobacco, nutrition)^{63,68,71–74}; and five studies addressed health information, education, or literacy.^{59,66,71,73,74} Over the past 25 years, individual interventions have varied as have the quantity of social needs that have been addressed (Figure 2).

FIGURE 2.

Bubble plot of ED care transition intervention and number of social needs addressed. CGA, comprehensive geriatric assessment; SDOH, social determinants of health

Outcomes of interventions addressing social needs

Of those care transition intervention studies specifically addressing social needs, primary outcomes frequently included subsequent health care utilization including ED recidivism, hospital admission, nursing home admission, and outpatient follow-up. Time frames for ED revisit assessment varied from 7 days to 3 months. Quality of life and functional dependency were also primary outcomes in two studies.^62,64 Secondary outcomes studied most commonly included ED revisitation or hospital admission, patient satisfaction, and quality of life or functional dependency. Additional secondary outcomes studied included mortality, economic analyses of the intervention, primary care follow-up, nursing home admission, and referral completion.

We identified mixed results when specifically assessing the four intervention studies^63,68,71,74 that addressed at least five of the individual social need domains. Miller et al.⁷¹ performed a non-RCT and identified no significant reduction in health care utilization after a geriatric nurse clinician assessment. After a CGA with particular focus on social need components, Mion et al.⁶³ identified lower nursing home admissions and increased patient satisfaction with ED discharge, but failed to affect overall ED or hospital service usage at 30 or 120 days. In a pre-post study by Hegney et al.,⁶⁸ risk assessment and community referrals by an in-ED community nurse resulted in a substantive absolute 16% (21%–5%) decrease in ED revisit rate in the 8 months following the intervention. After providing an in-ED CGA and in-person follow-up, Ballabio et al.⁷⁴ addressed several SDOH domains and identified reduced ED revisit rates and a statistically significant improvement in physical and emotional status.

Consensus of highest-yield ED care transitions future research priorities

Through a series of monthly conference calls, the GEAR CT subgroup initially developed and proposed 25 potential research priorities regarding care transitions. All GEAR Network members were invited to comment and provide feedback on the priorities over email. During the 2-day in-person GEAR Consensus Conference, a revised list of potential priorities were presented and discussed in small- and large-group forums. With guidance and prioritization by the GEAR CT subgroup, the group collectively agreed on the top five research areas to be ranked in order of prioritization. Assessing those five research areas by priority rank (1–5), there was 100% voting participation by all 49 members of the GEAR Consensus Conference, with stakeholder attendees obtaining agreement on the order of primary research topics to accelerate ED care transitions research (Table 1). “Determining optimal outcome measures for ED–home transition interventions” was identified as the highest priority area for future ED-based care transitions research.

TABLE 1.

Consensus conference ranking of ED care transition research priorities

Research priority rank—descriptor
1—Determining optimal outcome measures for ED-home transition interventions
2—Identifying optimal candidates for additional support during the ED-home transition
3—Improvement of bidirectional nursing home-ED transitions
4—Linkage of ED data to social risk factors, fall risk, physical function, and medication safety
5—Incorporation of stakeholder perspectives in ED care transitions research

DISCUSSION

This scoping review examines the available literature on ED-based care transition interventions for older adults that specifically consider social needs. Our analysis demonstrated significant diversity among the included studies in terms of study design, intervention types and components, social need domains addressed, and outcome measures. Overall, results from these studies demonstrated variable impact on health care utilization and other selected outcomes, and there was no consistent pattern between positive outcomes and the incorporation of a care transition intervention addressing a specific social need domain.

With older adults particularly susceptible to certain social risk factors, our work builds upon a recent systematic review of all adult ED patients that failed to demonstrate a reduction in repeat ED visits or hospital admission after care transition interventions.⁷⁵ Our work robustly assesses interventions from a wide number of international studies, allowing translational components to be considered from countries that have been noted to place a greater emphasis on social services.⁷⁶ This analysis also uniquely focuses on social needs and includes all possible care transition interventions, as opposed to reviews^46,48,55,56 previously focused solely on the impact of one intervention or assessment type. In comparison to SDOH-related reviews of the broader adult ED population,^22,35–37 our review specifically of the geriatric population identifies an encouraging proportion of care transition intervention studies addressed at least one social need component. Furthermore, this consensus conference approach builds upon and adds to the nearly two decades of Academic Emergency Medicine proceedings highlighting the need for multidisciplinary care in helping vulnerable populations, such as older adults impacted by social risk factors, manage the increasingly complex health care system.^77–79Health care systems are increasingly invested in improving care transitions as an important way to improve clinical outcomes and reduce fragmentation of care and preventable health utilization.^80,81 Clinicians are simultaneously incorporating new tools to address patients’ social needs in the clinical setting itself through a variety of mechanisms, including new technology to capture social data and the use of more social services staff in the clinical setting.^82–84 The development of new care transition interventions should take advantage of these novel tools, as there is significant room for clinical improvement, evidenced recently by Benjenk et al.⁸⁵ noting the mixed rates of adherence to discharge instructions and the potential need for additional care transition support for older adults discharged home from the ED. A more rigorous and systematic approach to incorporating new data and tools that address social needs holds significant promise, given its increasing emphasis in general clinical care delivery.

Our work presents a number of findings to guide best research practices going forward. We note that 14 of the 17 identified studies addressed at least one social need. It is evident that researchers in this field recognize the impact of social risk factors in ED care transitions for older adults, similar to those in other settings.^12,86 However, a lack of common terminology exists regarding SDOH care transition interventions within included studies. This lack of common terminology makes it difficult to understand how much intention was given toward the inclusion of a social need component or if it was largely incidental within what was intended to be a medically oriented intervention. Also, this lack of common terminology around SDOH is not unique to the care transition intervention literature and additionally makes it difficult for researchers to communicate their findings to other stakeholders like policymakers, the social services sector, and payers.¹⁷ To move the field forward, we agree with the recommendation¹⁶ to use a common agreed-upon language among researchers when describing “social determinants of health,” “social risk factors,” and “social needs.”

The results of this review underscore the recommendation of the GEAR Consensus Conference that determining optimal outcome measures be considered a high priority to accelerate ED care transitions research. Specific care transition outcome measures previously studied in the ED have been limited by their attempted translation from the inpatient hospital setting or absence of dedicated focus to the geriatric population.⁸⁷ Historically, outcome measures used for interventions that address social or economic determinants of health have focused primarily on process and social outcomes and less frequently on health care utilization.⁸⁸ Previous recommendations in the geriatric emergency medicine literature have centered on care transitions between nursing homes and EDs, but are often applied broadly to other care transitions.⁸⁹ Collectively, the prioritization by the GEAR Network of determining optimal outcome measures for older adults experiencing care transitions supported by the limited available patient-reported outcome measures highlights a critical knowledge gap that researchers should address.

The focus on social needs also has relevant policy implications. Through accountable care organizations (ACOs) and other value-based care models, greater attention has been directed toward promoting and paying for social risk screening.⁹⁰ For Medicare beneficiaries potentially impacted by the findings of this analysis, the Centers for Medicare & Medicaid Services recently initiated a program, titled the Accountable Health Communities Model, to address the health-related social needs through enhanced clinical-community linkages.⁹¹ Furthermore, in the shift toward value-based payments, the increased adjustment for social risk factors and greater investment in upstream mechanisms of social support are only projected to increase.^92,93 EDs will continue to play a significant role in the coordination of care for older adults, and although they have not systematically been integrated into value-based payment reforms, they will need to develop better mechanisms for coordination with primary care providers and specialists within ACOs and similar groups.⁹⁴ Improving ED-based care transitions will likely be a significant part of care coordination within any given network, and so new interventions in this area are essential to the goals of value-based payment reform. Other payment incentives may need to be implemented to promote the use of these care transition interventions and any technological or social services that prove to be effective elements of an intervention.

LIMITATIONS

Our findings should be considered in light of several limitations. First, we may not have captured all relevant articles based on our search strategy. To manage a relatively large and disparate body of literature we adopted a pragmatic approach⁹⁵ and relied on existing systematic reviews within phase I. We included as many relevant key terms as possible within the search strategy and ensured, with the assistance of a medical librarian, that the search strategy identified exemplary articles for the PICO question. Second, given the purpose of a scoping review, we did not perform risk-of-bias or quality appraisals of the included studies. To capture the breadth of available care transition interventions focusing on social needs in older adults, we chose to search for a wide range of interventions and outcomes and therefore could only conduct an exploratory assessment of patterns of effectiveness. We also recognize that various definitions of SDOH exist. In using the PRAPARE tool framework, we may have neglected to include certain SDOH domains that some may view integral to older adult care transitions. Finally, classifying interventions using other frameworks, such as the 5As framework,^96,97 may yield complementary or additional insights regarding ED-based care transitions interventions addressing social needs.

CONCLUSIONS

This scoping review identifies that ED care transition interventions in older adults addressing social needs frequently use multicomponent interventions. Studies often address at least one social need component, yet exhibit variation in their degree of successful impact on a wide array of health care utilization and patient-centered outcomes. Identification of optimal outcome measure for ED–home transition interventions was identified at the Geriatric Emergency care Applied Research Consensus Conference as the highest priority area for future ED-based care transitions research.