Involving Nursing Home Residents and Their Families in Acute Care Transfer Decisions

Involving Nursing Home Residents and Their Families in Acute Transfer Decisions: Pilot Test of a New Decision Aid

This study addresses the Patient-Centered Outcomes Research Institute (PCORI) area of interest related to development of decision support tools that bring patients, families and clinicians together to decide, in this instance, whether or not transfer from the nursing home (NH) to acute care is necessary and appropriate. The purposes of this study were 1) to develop an evidence-based decision aid addressing potentially avoidable transfers of residents from nursing homes to hospitals (preceded this protocol), and 2) to evaluate this decision aid in terms of acceptability to residents and families and its effect on the quality of transfer decisions. The primary hypotheses to be tested are: Hypothesis 1: Resident and family members in the intervention group will report greater preparation for decision making and less decisional conflict than those in the no treatment control group. Hypothesis 2: Residents and family members in the intervention group will demonstrate increased knowledge related to acute care transfer and less preference for acute care transfer than those in the no treatment control group.

Project Summary: The goal of this project is to develop (Phase I) and test (Phase II) a new Transfer Decision aid that will better prepare nursing home (NH) residents and their families to participate in decisions to transfer the resident to an acute care facility or remain in the NH for treatment. Objectives: The specific aim of this study is to evaluate the decision aid, "Go to the Hospital or Stay Here?" in terms of acceptability to residents and families and its effect on decision making and transfer decisions by residents and families. Study Design: A mixed methods design was used to address the study aims. In Phase I (not submitted to ClinicalTrials.gov), interviews of a diverse sample of NH residents, their family members, primary care providers, nurses and social workers in the NH were conducted and the results were used to create a Transfer Decision aid to be pilot tested in Phase II for its effect on decision making and transfer decisions. Phase II involves a pretest, intervention for the half randomly selected to receive it, 14 day posttest and 90 day follow-up. Study Population: Staff of 15 participating South Florida nursing homes (NHs) were asked to recommend residents and family members of residents who could participate in the study. Residents were tested for ability to provide consent using the Mini-Cog prior to enrollment. General Analytic Strategy: Multiple linear regression (MLR) will be used to analyze the variance when predicting the primary and secondary outcome variables, preparation for decision making, decisional conflict, knowledge and number of transfers (McNeil, Newman & Kellee, 1996; Pedhazur & Schmelking, 1991). For qualitative data, Phase II interviews are transcribed and the transcriptions reviewed for accuracy. Responses were de-identified (names, gender and ethnic identity information removed). A framework for descriptive (labeling and categorizing), interpretive (based on underlying meaning) and pattern (thematic) codes was developed. Intercoder reliability was calculated.

Residents and families randomly assigned to the intervention group (one half of the residents and one half of the families enrolled) are given the new Decision Aid to review with an RA trained for this purpose by the investigators. The Decision Aid provides information on risks and benefits of acute care transfer and information on advance care planning, resident and families' right to be involved in the decision. Resident or family member is also asked to re-read it and think about it over the subsequent 14 days.

Range of possible raw scores: 10 - 50. Items can be summed and scored (sum the 10 items and divide by 10). There are no subscales. A higher score is positive indicating perception of being better prepared to make the decision identified when test is administered.

Decision Aid group only: Post test (2-3 weeks after pretest) and at 3 month follow-up (after post test)

Scale Range of possible raw scores: 0 - 64. A lower score indicates less decisional conflict related to making the identified decision. Decisional conflict is a state of uncertainty about a course of action. Decision supporting interventions are thought to be effective in

Decision aid: Pretest (baseline) and Post test (2-3 weeks after pretest); No decision aid: Post test only

Decision Aid Quiz Investigator Developed: True/False Test of Knowledge based upon information in the Decision Aid. Scale Range: 0 - 22. Total score is the number of questions answered correct. A higher score indicates greater number of responses are correct. Items are scored as 0 = incorrect; 1 = correct. Criteria used to assess this measure: Items based upon concept mapping of the Decision Aid Reviewed by a panel of experts prior to administration

Treatment Site Option Preference Administered prior to the Decisional Conflict Scale, this single item was used to ask participants if they preferred treatment in the hospital or nursing home or if the preference depended upon the seriousness of their condition (three choices). This is nominal level data. The administration of the item precedes administration of the Decisional Conflict Scale (DCS) described previously. Standard administration of the Decision Conflict Scale includes this item but it is not part of the DCS score.

Report of all hospitalizations from facility staff or participants Hospital Transfers Occurring During Study Period, i.e., number of resident participants who were hospitalized from pretest to posttest and by end of 3 month follow-up period.

Inclusion Criteria: Adult nursing home resident and/or family member or identified significant other of a nursing home resident. Cognitively unimpaired individuals. Long or short term residents of nursing home or rehabilitation center. Exclusion Criteria: Dementia as indicated by score on Mini-Cog. Inability to respond to questions due to physical disability or illness.

Tappen RM, Worch SM, Elkins D, Hain DJ, Moffa CM, Sullivan G. Remaining in the nursing home versus transfer to acute care: resident, family, and staff preferences. J Gerontol Nurs. 2014 Oct;40(10):48-57. doi: 10.3928/00989134-20140807-01.