Pilot Project: Helping Nursing Home Residents and Their Families Decide Whether to Go to a Hospital for a Medical Problem

PCORI funded the Pilot Projects to explore how to conduct and use patient-centered outcomes research in ways that can better serve patients and the healthcare community. Learn more.

Background

Almost 25 percent of nursing home residents go to the hospital for medical treatment within the first month of living in their nursing home. These hospital trips cost Americans $14.3 billion each year. Nursing homes residents who go to the hospital might get sicker while there. Studies suggest that two-thirds of these hospitalizations could be prevented. Many times, nursing home residents go to the hospital because they or their families insist on it.

Project Purpose

The research team wanted to understand what nursing home residents and their families need to know when deciding whether to go to the hospital. Researchers wanted to learn how residents and families want to be involved in this decision and what factors are most important to them. The research team also wanted to determine whether a decision aid affected residents’ and families’ decision-making process about whether to go to the hospital. Decision aids help people choose between two or more healthcare options based on what is most important to them.

Methods

This study had two parts. First, researchers interviewed 96 nursing home residents, 75 family members, and 100 healthcare providers at 19 nursing homes. Researchers asked about residents’ and family members’ experiences with deciding whether to go to the hospital for a medical problem or to stay and get treatment in the nursing home. Researchers used the information from the interviews to create a decision aid called Go to the Hospital or Stay Here?

For the second part of the study, researchers assigned 191 nursing home residents from 15 nursing homes to two groups. In one group, 96 nursing home residents received the decision aid. The other group of 95 residents did not receive the decision aid. Three months later, researchers used a knowledge test to compare the two groups’ knowledge about when to go to the hospital, feeling of preparation for making decisions, and number of transfers to the hospital. Researchers also looked for patterns in residents’ preferences about getting treatment in the nursing home or in the hospital.

Findings

From the interviews, the research team found that, before participating in the study, most nursing home residents had never thought about whether it was better to go to the hospital or stay in the nursing home for treatment. About half of family members and nursing home staff also had never thought about this question before.

Family members wanted more information overall. Nursing home residents wanted more involvement in deciding whether to go to the hospital. Residents were more likely than family members to say their choice to go to hospital would depend on how severe their health condition was.

Residents and families described their nursing home and hospital experiences in detail. They understood the risks and benefits of staying in the nursing home versus going to the hospital. For example, they worried that nursing homes did not have enough staff to care for sick residents, but they liked that the staff know residents personally. They thought staff at the hospital might not understand residents as well as the nursing home staff did.

Most research on decision aids focuses on how patients feel about the decision-making process and how they use information to make decisions. In the first part of the study, researchers found a third consideration: Some residents said they had asked others to make the decision for them. They did so because they did not feel like they could make an important decision on their own or that they were too frail, ill, or tired.

In the second phase of the study, nursing home residents who received the decision aid went to the hospital about as often as residents who did not get the decision aid. Residents who received the decision aid had significantly more knowledge about the decision to go to the hospital than residents who did not receive the decision aid. Almost all the participants who received the decision aid said it was helpful.

Limitations

Many nursing home residents did not complete the second phase of the study, and researchers could not follow their progress to find out whether they took additional trips to the hospital. Results might not be the same with a larger group of residents who all complete the study.

Conclusions

Deciding whether to go to the hospital is a complex process for nursing home residents and their families. Researchers found that most participants had never thought about having to make this decision. Participants found the decision aid helpful, but it did not affect decisions about whether or not residents would go to the hospital.

PCORI funded the Pilot Projects to explore how to conduct and use patient-centered outcomes research in ways that can better serve patients and the healthcare community. Learn more.

Project Purpose

Almost one in four individuals admitted to a nursing home (NH) is hospitalized within 30 days. Such hospitalizations cost about $14.3 billion annually in the United States. Studies suggest that up to two-thirds of such hospitalizations may be preventable. Hospitalization carries risks including delirium, falls, pressure ulcers, hospital-acquired infections, and a decline in function. An intractable reason for potentially preventable hospitalizations is the insistence of the NH resident or a family member on such a transfer.

Research Questions

This study sought to answer the following research questions: (1) What are NH residents’ and families’ information needs, preferred degrees of involvement, and associated values related to the decision to either transfer a resident to the hospital or have them treated in the NH? (2) What effect does a newly developed resident and family guide (decision aid) about hospital transfer have on their knowledge, preparation, decisional conflict, and actual hospital transfer?

Study Design

This was a two-phase, mixed-methods study. In Phase I, researchers interviewed 271 NH residents, families, and providers and developed a family decision aid (“Go to the Hospital or Stay Here?”) in both English and Spanish. Phase II was a pilot test of the guide.

Participants, Interventions, Settings, and Outcomes

Phase I included 96 NH residents, 100 staff, and 75 family members drawn from 19 nursing homes. Phase II included 96 intervention and 95 control patients drawn from 15 nursing homes.

Phase II intervention participants received the decision aid; control participants did not. Outcome measures included scales for knowledge, preparation for decision making, NH versus hospital preference, and hospital readmissions.

Data Sources

Phase II included NH records.

Data Analysis

In Phase I, researchers conducted open coding of interview transcripts to identify themes. The research team reviewed codes and applied them to each interview. Intercoder agreement was 87.5%. The research team merged coded responses with sociodemographic data and performed logistic regression to identify characteristics predicting preference for treatment in the NH or hospital transfer. A qualitative approach was used to analyze residents’ and family members’ descriptions of experiences in the hospital and NH. Frequencies were reported. Researchers performed cognitive task analysis to identify the decision mode used and compared modes used by residents and family (chi-square tests).

In Phase II, researchers used hierarchical linear mixed-model analysis to evaluate differences in knowledge from pretest to posttest across groups. They employed ANOVAs to evaluate decisional conflict, preparation for decision making, hospital transfers, and transfer-related preferences. Frequencies and chi-square tests were used to examine differences in preferences for treatment in the NH or transfer to the hospital.

Findings

Most residents and half of the family members and staff had not given any thought to the first research question (regarding information needs, preferred degrees of involvement, and associated values related to the decision to either transfer a resident to the hospital or have them treated in the NH) before the Phase I interview. Family members wanted more information while residents thought they should be more fully involved in the transfer decision. More residents than family members indicated that their preference would depend on the severity of their condition and their prognosis. Results suggested that discussion of transfer decisions should occur soon after admission and that differences in perspectives of residents and family members may be expected.

Residents and family members provided vivid descriptions of their NH and hospital experiences. This suggested that they are aware of the risks of both options; of particular concern was the limited resources and staffing in many NHs and the impersonal care received in hospitals. A critical reason for preferring to remain in the NH if possible was that, “they know me here.”

Most studies of decision-making modes identify affective and information-seeking modes. Researchers in this study identified a third mode, namely delegation to trusted others, which was used primarily by residents. Researchers identified several reasons that residents were more likely than family members to delegate the transfer decision to others. These included fatigue, illness, or frailty; uncertainty regarding their own ability to make an important decision; and feeling of vulnerability related to the consequences of the decision.

In Phase II the researchers found a statistically significant increase in knowledge related to this important decision in the intervention group. Hospital readmissions and decisional conflict were not significantly reduced. Those who received the decision aid (guide) were overwhelmingly positive in evaluating its helpfulness; 93% rated it as helpful or very helpful.

Limitations

The primary limitation occurred in Phase II as a substantial number of NH residents were lost to 3-month follow-up.

Conclusions

The reasons for resident or family member insistence on hospital transfer are complex, but several critical points for intervention exist to help with the decision. A decision aid such as the one used in this study provides an opportunity to think about this issue before a crisis occurs. Involving residents and families in the decision itself does not increase insistence on hospital transfer. Nursing homes can also do a great deal to increase resident and family trust in the ability of NH staff to provide adequate monitoring and treatment.