Depression has pervasive effects on quality of life and health outcomes for elderly nursing home (NH) residents, but existing care processes are poorly understood and the knowledge base for improvement is inadequate. This study will provide a comprehensive assessment of patterns of depression identification and treatment for elderly residents, using merged data from the Minimum Data Set (MDS), Medicare and Medicaid claims, the OSCAR database on facility characteristics, and information on community characteristics in 6 states. The research dataset will integrate extensive clinical information from MDS with details on health and mental health services use, diagnoses, and pharmaceutical use histories from claims, along with contextual information on facilities that will relate care of individuals to other aspects of nursing facility care and life. The study aims to provide a clearer understanding of variations across facilities and communities in depression-related care processes, and to articulate the effects of facility structure, resources, case-mix, and practice patterns on resident care and outcomes. It will examine patterns of identified depression symptoms and diagnoses; antidepressant treatment; use of mental health services; and their interrelationships. An important focus will be on the analysis of racial/ethnic and other socioeconomic and geographical disparities, and identification of policy-modifiable and other factors that give rise to them. Trends in depression identification and treatment from 1999 through 2007 will be evaluated, including changes following public release via the Internet of a facility-level depression quality measure by CMS in 2004. Longitudinal analyses will evaluate trajectories of depression-related resident outcomes over time. Guided by a conceptual framework that articulates hypothesized processes leading to depression identification, treatment, and outcomes in a process of interaction among resident, facility, and community factors, the study will use multilevel modeling techniques that account for clustering of repeated observations within residents, residents within facilities, and facilities within communities.
Showing the most recent 10 out of 16 publications
