Validation of the Geriatric Depression Scale for an elderly Sri Lankan clinic population.

Background: Geriatric Depression Scale (GDS) has not been validated for the elderly population in Sri Lanka. Aim: To translate, validate, and examine the effectiveness of GDS and to suggest the optimal cut-off scores for elderly Sri Lankans attending a psychogeriatric clinic. Materials and Methods: The Sinhalese translation of GDS (GDS-S) was administered to people aged 55 years and above, attending a psychogeriatric outpatient clinic. The diagnostic performance of the instrument was compared against the ICD 10 diagnosis of a consultant psychiatrist, which was considered the 'gold standard'. Receiver operating characteristic (ROC) analysis was carried out to compare the diagnostic performance of the GDS-S. Optimal cut-off scores for depression and sensitivity and the specificity of the instrument was determined. Results: A total of 60 subjects formed the final sample (male/female=16/44) of which 30 were depressed, while 30 were age- and sex-matched controls. The optimal cut-off score for GDS-S was 8 for differentiating non-depressed from mildly depressed, while the cut-off score for moderate depression was 10. Sensitivity and specificity of GDS-S was 73.3% for differentiating depressed from non-depressed. Conclusion: GDS is culturally acceptable, easy to use, sensitive, and a valid instrument to diagnose depression and to differentiate mild from moderate depression in an elderly Sri Lankan clinic population. [ABSTRACT FROM AUTHOR]