Performance of case-finding tools for depression in the nursing home: influence of clinical and functional characteristics and selection of optimal threshold scores.

OBJECTIVE: To compare case-finding tools for depression in the nursing home setting and to evaluate effects of subject function, cognition, and disease number on test performance. DESIGN: Cross-sectional survey. SETTING: One academic and four community homes. SUBJECTS: One hundred thirty-four randomly selected, mildly cognitively impaired, functionally dependent residents. METHODS AND MEASURES: The Geriatric Depression Scale (GDS), Short Geriatric Depression Scale (SGDS), Center for Epidemiologic Studies Depression Scale (CES-D), and Brief Carrol Depression Rating Scale (BCDRS) were administered. The Structured Clinical Interview for DSM-III-R diagnoses was administered independently,. Operating characteristics and the effects of subject characteristics on test performance were evaluated using McNemar's test and logistic regression. Selection of "optimal" threshold scores was guided by Kraemer's quality indices and clinical judgment. RESULTS: Thirty-five subjects (26%) had major depression. No differences were found among the instruments in sensitivity (range 0.74-0.89), specificity (range 0.62-0.77), or area under the receiver operating curve (ROC) (range 0.85-0.91). Resident characteristics did not affect test performance. Quality indices showed the GDS and BCDRS met criteria for moderate to substantial agreement with the criterion standard, whereas the SGDS and the CES-D achieved only fair agreement. No change in threshold scores was warranted. CONCLUSIONS: The GDS and BCDRS performed well in the nursing home. As the GDS can serve as a both a case-finding and severity instrument, it is preferred. Use of brief, interviewer-administered tools may improve detection of depression in the nursing home.

Duke Scholars

Author John Wiley Williams Jr. Medicine, General Internal Medicine