RESUMO
BACKGROUND: Research on end-of-life care in nursing homes is hampered by challenges in retaining facilities in samples through study completion. Large-scale longitudinal studies in which data are collected on-site can be particularly challenging. OBJECTIVES: To compare characteristics of nursing homes that dropped from the study to those that completed the study. METHODS: One hundred two nursing homes in a large geographic 2-state area were enrolled in a prospective study of end-of-life care of residents who died in the facility. The focus of the study was the relationship of staff communication, teamwork, and palliative/end-of-life care practices to symptom distress and other care outcomes as perceived by family members. Data were collected from public data bases of nursing homes, clinical staff on site at each facility at 2 points in time, and from decedents' family members in a telephone interview. RESULTS: Seventeen of the 102 nursing homes dropped from the study before completion. These non-completer facilities had significantly more deficiencies and a higher rate of turnover of key personnel compared to completer facilities. A few facilities with a profile typical of non-completers actually did complete the study after an extraordinary investment of retention effort by the research team. CONCLUSION: Nursing homes with a high rate of deficiencies and turnover have much to contribute to the goal of improving end-of-life care, and their loss to study is a significant sampling challenge. Investigators should be prepared to invest extra resources to maximize retention.