INTRODUCTION

The number of older adults transitioning to skilled nursing facilities (SNFs) is rapidly increasing,1 and the aging of the U.S. population suggests this trend may only accelerate. Emerging payment reforms are sharpening focus on patient selection for SNF and on outcomes such as readmissions, rates of discharge back to the community, and costs.2,3

Despite the national focus on post-acute care brought about by legislation, relatively little is known about how hospitalized older adults and their caregivers decide whether to go to a SNF for post-acute care. Published studies have focused on the content of the information presented, rather than trying to assess the decision-making process4,5 or on caregivers who recently placed a family member in a long-term care facility from home, rather than inpatients going to post-acute care.6

We sought to understand to what extent hospitalized older adults and their caregivers are enabled to make a high-quality decision about post-acute care in an SNF. We interviewed patients and caregivers in both the hospital and SNF setting about their experience with decision-making around SNF placement to elicit elements of context or process influencing their satisfaction with the outcome of this decision. This information is crucial to inform future efforts to align patient needs and goals with resources following hospitalization.

METHODS

Study Design

This analysis is part of a larger qualitative study evaluating patient and provider decision-making regarding post-acute care in SNFs from both the hospital and SNF perspectives. We identified an evidence-based conceptual framework that describes key components of high-quality decision-making—the Ottawa Decision-Support Framework (ODSF)7—and used a framework design to guide our analysis.8 The ODSF draws on concepts from psychology, decision analysis and conflict, social supports, and economic theory to assess patient needs for making a high-quality decision, what supports are available, and evaluation of the outcomes of the decision made.9 It has been widely used to support development of patient decision aids in clinical situations where achieving a high-quality decision is challenging.10,11,12,13,14

Setting and Participants

We identified participants in three hospitals and three SNFs, including several different clinical units within the hospitals. The three participating hospitals included a VA hospital serving a predominantly male population often with significant medical comorbidity and weak social supports,15,16 a quaternary-level university hospital serving a large referral base as well as a predominantly urban population, and a safety-net public hospital serving a predominantly indigent and immigrant population. We sampled participants from clinical units that primarily discharge older adults to SNFs, such as general medical wards, an Acute Care for the Elderly unit,17 and an orthopedic surgery unit. SNFs sampled included: a VA Community Living Center (CLC) co-located with the main VA hospital that provided short-term rehabilitation only; a predominantly long-term, Medicaid-funded nursing home with a smaller Medicare-certified rehab unit; and a community SNF that only provided short-term rehabilitation under Medicare. The requirement for written informed consent was waived; the study was approved by the Colorado Multiple Institutional Review Board.

Hospitalized patients were eligible for inclusion if they had an unplanned hospitalization, were over age 65, and were being discharged to an SNF. We identified eligible patients through participation in interdisciplinary rounds on each unit or by hospital staff referral. Study staff relied on the expertise of the treating teams to determine if a patient was cognitively able to participate in an interview; patients deemed unable were excluded. SNF patients were eligible if they had recently been discharged from an acute care hospital for short-term rehabilitation, were over age 65, and did not have a level of cognitive impairment that would prevent them from meaningfully participating in the interview (determined by primary treating physician). We relied on referrals from Medical Directors, attending physicians, and Directors of Nursing at each SNF to identify eligible patients. We identified caregivers by asking patients, “Can you think of the one person besides a health care provider who helps you the most with your medical care?”18

Data Collection and Analysis

Between February and September 2016, qualitative analysts (EL, AL, RA) conducted a 20–60-min semi-structured in-depth interview with each participant. Interviews with patients were conducted in person, while interviews with caregivers were conducted in person or over the phone. Patients and caregivers were interviewed separately. We used an interview guide informed by prior research, theories and frameworks related to transitions of care and clinical experience (see Online Supplementary Material).19,20,21,22 Topics included: evaluating the need for post-acute care; selecting post-acute care options; attitudes, knowledge, and beliefs about SNFs; influences on discharge decision-making; and post-discharge follow-up. Interviews were audio-recorded, professionally transcribed, validated, and analyzed in Atlas.Ti (v7.5.11; Scientific Software Development, Berlin, Germany). Participant demographics were collected using a brief questionnaire before each interview.

We employed a team-based approach to the framework analysis method. We initially used a deductive approach, looking for key components of the ODSF that are key components of high-quality decision-making. However, we discovered many key components of the ODSF were not discussed by participants, and the process did not fit this ideal decision-making model. We thus re-approached the data using an iterative inductive-deductive approach, identifying key themes and grouping them into broader concepts from the ODSF (context, process, and outcome). We used a framework matrix to identify patterns in the context and process characteristics for patients and caregivers (separately) where the outcome was optimal as described by the ODSF: informed, aligned with values, and not associated with regret or blame.7 Finally, we identified suggestions of patients and caregivers for improving the decision-making process.

Initial codes were developed through group discussion; additional codes were developed in the process of thorough re-immersion in the transcripts by individual team members. As new codes emerged, they were discussed at team meetings to reach consensus on code labels and definitions until saturation was reached.23 To ensure reliability of our findings we (EL, AL, RA, CL, RB) reread selected transcripts to confirm themes and patterns identified. We met weekly throughout data analysis to discuss the process and emergent themes, and key analytic decisions were documented as part of our audit trail.

RESULTS

We interviewed 32 patients and 22 caregivers (n = 54 total), across hospitals (n = 32) and SNFs (n = 22). Patients were generally elderly and white with Medicare as their payer. More than 80% identified a caregiver, half of whom were children of the patient (Table 1). Overall, our results suggested most patients experience significant challenges to making a high-quality decision about post-acute care in a skilled nursing facility related to both contextual and process factors, resulting in suboptimal outcomes of their decisions.

Table 1 Patient and Caregiver Participant Characteristics

Decision Context

The ODSF recognizes that high-quality decisions are affected by the personal and clinical characteristics of the person making the decision, as well as key elements of the external context in which the decision is being made.7

Patients and caregivers described barriers across these contextual domains (clinical, personal, and external) to high-quality decision-making. The presence of active medical problems that impaired full engagement in the decision being made was the main clinical barrier cited by patients and caregivers. Their prior experience failing at home and being readmitted, experience with SNF, sense of identity, and their social support were predominant personal drivers of decision-making. The pressure to quickly make a decision participants felt unprepared for was the main external contextual influence.

Patients described a variety of clinical reasons for difficulty attending to a decision about post-acute care, including receiving sedating medications, having their sleep/wake cycle disrupted, or having unresolved symptoms (such as weakness or pain; Table 2, quotes 1–3). Their prior personal experience with post-discharge care also strongly influenced their current decision-making process. For example, several patients had tried going home and been readmitted to the hospital or had been to SNF before, and these experiences were described as framing the decision to go to SNF after the current hospitalization (Table 2, quotes 4–6). The involvement of a caregiver (or lack thereof) was a key contextual determinant of SNF decision-making (Table 2, quotes 5–7). However, patients and caregivers experienced the decision as unexpected and rushed (Table 2, quote 8).

Table 2 Key Quotes Regarding Decision Context

Decision Process

A high-quality process as defined by the ODSF is one in which the patient’s knowledge, values, and preferred role in decision-making are elicited and their needs identified.7 Surprisingly, patients reported passive participation in the process and heavily relied on the recommendation of the hospital care team (Table 3, quotes 1–2). Very few described considering other post-discharge options and most described minimal discussion with the care team about the decision (Table 3, quote 3). In many cases, patients saw discharge to SNF as a way to leave the hospital, since they perceived their only other option was to remain in the hospital (Table 3, quote 4). Those who tried to understand how SNF placement might fit into their goals expressed frustration with how little information they were provided by the hospital care team (Table 3, quotes 5–6).

Table 3 Key Quotes Regarding Decision Process

Decision Outcome

An ideal decision outcome is one that was informed, aligned with the patient’s values, and not associated with regret or blame.7 Since patients and caregivers described significant challenges to being informed and had difficulty connecting their values to SNF stays, patients most commonly expressed resignation and lack of choice and autonomy (Table 4, quotes 1–2). The relatively passive engagement with the process was striking given participants expressed a significant concern about loss of autonomy, and many had to undergo the transition to SNF before understanding how it might align with their goals (Table 4, quote 3). When SNF care was unable to realize those goals, participants were dissatisfied with the choice (Table 4, quote 4).

Table 4 Key Quotes Regarding Decision Outcome

Key Ingredients for a Positive Decision Outcome

We identified few patients or caregivers who described their decision as informed, aligned with their values, and not associated with regret or blame. When this did occur, common themes that arose included their active engagement in the decision-making process (Table 4, quote 5), realization of how SNF is aligned with expressed care goals (Table 4, quote 6), and having both prior experience with SNF and control over which SNF they will go to (Table 4, quote 7).

Suggestions for Improvement

Patients expressed interest in being more active participants in decision-making, even though their ability to do so was often impaired because of sedation or acute medical illness. When discussing SNFs with patients, patients suggested two areas of improvement: first, they wanted to know what actually took place in an SNF and what it would be like on a day-to-day basis while they were there. Particularly for patients who saw this as a challenge to their identity as independent older adults, learning about and experiencing the therapeutic and recreational programs at the SNFs helped assuage their fears (Table 5). Second, they wanted unbiased reviews from multiple sources to assist in decision-making, from results of inspections, quality metrics, and being able to call a patient who had recently been there to hear about their experience. Caregivers strongly felt patients should have more time in the hospital prior to discharge to an SNF to be able to more fully recuperate and felt they should be much more involved in the decision-making, particularly given the frequent perceived inability of their loved ones to participate meaningfully in decision-making.

Table 5 Patient and Caregiver Suggestions for Improvement

DISCUSSION

Post-acute care payment reforms are placing the decisions older adults make about post-discharge supports at the center of a national debate about “who should go where.”24 Our results suggest hospitalized older adults and their caregivers are infrequently able to make a high-quality decision about post-acute care because of important contextual and process factors, leading to significant dissatisfaction with the ultimate outcome of the decision.

Perhaps the most significant result of our investigation is that supporting high-quality decision-making for patients regarding post-acute care requires far more than providing information. This has been the main paradigm in the limited literature evaluating how to improve patient decision-making regarding post-acute options, summarized as: if we could just provide better information, patient decisions and outcomes will improve.4,5,6 We found patients and caregivers did desire objective information from a variety of sources when making a decision, but that gaining information was only one aspect of a much larger decision-making process.

How can we then support high-quality decisions in hospitalized patients about post-acute care, including SNF? Using the principles of the ODSF and our results, we posit that the decision to pursue post-acute care in an SNF should be approached similarly to a “goals of care” conversation used in palliative discussions. Patients for whom SNF is recommended usually have multiple significant comorbidities and impaired functional status; in addition, more than half are 80 years of age or older.25 The stakes of such a decision are high, as costs of SNF care to patients can be significant, and failure to rehabilitate can lead to long-term nursing home placement.26,27 Attending to important contextual factors and using a structured process to elicit patient (and caregiver) values, goals, and preferred role in decision-making—to tailor how information is provided and recommend post-discharge options—holds the best chance of supporting a high-quality decision.

This work should be interpreted in the context from which it was derived. For example, our hospital interviews only took place in predominantly academic and tertiary centers and all interviews took place in a single urban area; our findings may not be generalizable to community or rural settings. Strengths include a large, diverse sample of patients and caregivers and the ability to capture decision context, process, and outcomes by conducting interviews in the hospital and SNF setting. We used a validated, widely used framework for our qualitative framework analysis and robust methods to assure analytic quality.

These results, including common patterns leading to a high-quality decision and suggestions for improvement, suggest the utility of a structured, patient- and provider-“facing” intervention to improve the quality of post-acute care decisions. This is critically important as the number of older adults being discharged to post-acute care supports continues to increase1,28 and incentives are rapidly changing as the result of post-acute care reforms.