Objective: To validate the short Geriatric Depression Scale (GDS15) as a screening instrument and determine the optimal cutpoint for detecting depression among older people living in the community.
Design: Two stage study with the first stage consisting of a health check of people aged 75 years and over by a practice nurse which included the GDS15. A second-stage diagnostic interview including the criterion standard was carried out blind to GDS15 score.
Setting: A large general practice in Melton Mowbray, Leicestershire, UK.
Patients: A random sample (stratified according to first-stage GDS15 score) of 257 older people living in the community, without significant cognitive impairment, were selected for the second-stage diagnostic interview.
Measures: The first-stage GDS15 score was compared with diagnostic status for depression and anxiety disorders (according to ICD-10 criteria) and presence of depressive symptoms at the second-stage clinical interview.
Results: Ninety-three per cent of those eligible for the study were successfully screened with the GDS15. A cutpoint of <3/3+ had a sensitivity of 100% and specificity of 72% in detecting cases of depression but fewer than one in five of those testing positive reached a diagnostic threshold. Only 25% of those with a diagnosis of depression had any mention of mental health problems in their medical notes in the year prior to the clinical interview. For detecting depressive symptoms the same cutpoint was 79% sensitive and 78% specific with a positive predictive value of 46%.
Conclusions: Used as part of an annual over-75 health check in primary care, the GDS15 would detect a significant amount of hidden morbidity which would serve the original purpose of the annual elderly health checks in the UK.
Copyright 1999 John Wiley & Sons, Ltd.