Prevalence and correlates of recognized depression in U. S. nursing homes

OBJECTIVES: To provide descriptive epidemiological information on identified depression in nursing home residents. DESIGN: Survey of the 1996 Medical Expenditure Panel Survey-Nursing Home Component (MEPS-NHC). SETTING: Nine hundred fifty-one nursing facilities. PARTICIPANTS: Three thousand seven hundred ten non-comatose residents. MEASUREMENTS: Demographic characteristics and health care and functional data were collected using key informant interview and medical-record review. Depression diagnoses were derived from Minimum Data Set (MDS) diagnoses and are nonspecific with regard to formal diagnostic criteria. RESULTS: The prevalence of identified depression was 20.3 cases per 100 residents (95% confidence interval = 5 18.9-21.7). Younger age, female sex, marital status other than never married, white non-Hispanic ethnicity, better cognitive functioning, heart disease, Parkinson's disease, and length of stay between 1 and 2 years were significantly and independently associated with a greater prevalence of identified depression. CONCLUSION: The prevalence of identified depression in the MEPS-NHC is lower than that previously estimated using formal diagnostic assessment techniques for threshold and subthreshold depressive states combined. Therefore, MDS-identified depression may underestimate the burden of depression. Underrecognition may be particularly acute in black or African-American residents, the oldest old, and the cognitively impaired. Further research on the relationship of MDS-identified depression and depression identified through structured diagnostic interviews in broadly representative samples of nursing home residents is needed to expand the availability of descriptive epidemiological data, to help clarify the results of research making use of MDS data, and to suggest methods for optimizing clinical and administrative data systems.