RESUMEN
OBJECTIVES:The purpose of this study is to evaluate the diagnostic validity of the three published short Geriatric Depression Scales and to identify a valid and reliable alternative to the Korean version of the Geriatric Depression Scale (GDS-K) and short form Geriatric Depression Scales (SGDS-K). METHODS:All subjects were registered elderly psychiatric patients who first visited the Geriatric Psychiatry Clinic of Seoul National University Hospital in Seoul between April 1996 and May 1999. The GDS-K, SGDS-K, Hoyl GDS-5, D'Ath GDS-4, van Marwijk GDS-4 were administered. Receiver operating characteristics (ROC) curve analysis performed to investigate the diagnostic validity of the Hoyl GDS-5, D'Ath GDS-4, and van Marwijk GDS-4. RESULTS:The best trade-off between sensitivity and specificity was achieved with a Hoyl GDS-5 score of 3 (sensitivity:0.7258, specificity:0.7717), a D'Ath GDS-4 score of 3 (sensitivity:0.5968, specificity:0.7391), and a van Marwijk GDS-4 score of 3 (sensitivity:0.6613, specificity:0.7391). The area under the ROC curves of each diagnostic tool were calculated into the Hoyl GDS-5 of 0.8198, the D'Ath GDS-4 of 0.7525, and the van Marwijk GDS-4 of 0.7528. CONCLUSION:The Hoyl GDS-5 may be utilized in screening for major depression among the elderly psychiatric patients in Korea.