An empirical typology of residential care/assisted living based on a four-state study.

Purpose: Residential care/assisted living describes diverse facilities providing non-nursing home care to a heterogeneous group of primarily elderly residents. This article derives typologies of assisted living based on theoretically and practically grounded evidence.
Design and Methods: We obtained data from the Collaborative Studies of Long-Term Care, which examined 193 assisted living facilities in four states: Florida, Maryland, New Jersey, and North Carolina. By using mixture modeling, we derived typologies in five different ways, based on: structure; process; resident case-mix; structure and process; and structure, process, and resident case-mix.
Results: Although configurations of typologies varied as a function of criterion variables used, common themes emerged from different cluster solutions. A typology based on resident case-mix yielded a five-cluster solution, whereas a typology based on structure, process, and resident case-mix resulted in six distinct clusters. Medicaid case-mix/psychiatric illness and high resident impairment were two clusters identified by both strategies.
Implications: Because of the wide variation in structure, residents, and services within assisted living facilities, typologies such as those described here may be useful in clinical practice, research, and policy. To the extent that public payment defines its own cluster, the potential for inequities in care merits careful attention.