Abstract
Anliegen Untersuchung der Validität der Geriatrischen Depressionsskala in der 15-Item-Version (GDS-15) sowie der Kurzformen GDS-8 und GDS-4 an einer Stichprobe von 92 Altenheimbewohnern. Methode Ermittlung der Validitätskennwerte bezüglich majorer und minorer Depressionen nach den Kriterien des Strukturierten Klinischen Interviews für DSM-IV (SKID) als Goldstandard. Ergebnisse Die diagnostische Güte der GDS-15 war höher für majore Depressionen als für minore Depressionen (area under the ROC curve: 89,7 vs. 73,4 %). Die GDS-8 erreichte vergleichbare Werte, die GDS-4 nicht. Schlussfolgerungen Die GDS-8 ist aufgrund ihrer Kürze der GDS-15 vorzuziehen. Die GDS-4 kann nicht empfohlen werden. Objective To assess and compare the validity of the German 15-item version and shortened versions of the Geriatric Depression Scale (GDS) for early detection of depression in nursing home residents with mild to moderate cognitive impairment (MMSE ≥ 15). Methods Data from 92 patients were analysed. The Structural Clinical Interview for DSM-IV (SCID) was used as the gold standard. The diagnoses covered in this assessment were current major depression (MD) and minor depression (MinD). The performance of the GDS scales was evaluated using receiver operating characteristics (ROC). Main outcome measures were AUC (area under curve) values, as well as sensitivity and specificity. Results Overall diagnostic validity of the GDS-15 was better for MD than for MinD (AUC: 89.7 % and 73.4 %, respectively). In assessing MD, AUC values as well as sensitivity and specificity were comparable for the GDS-15 and the GDS-8. A cut-off point ≥ 5 on the GDS-8 gave optimum performance (sensitivity: 76.9 %, specificity: 88.6 %). For the best cut-off point ≥ 2, the corresponding values of the GDS-4 were 53.8 % and 92.4 %. Conclusions The GDS-8 appears to be a less time-consuming alternative for the nursing home setting. However, the GDS-4 is not suitable for this population.