Background: depression is common in people with poor physical health, particularly within the acute medical in-patient setting. Co-morbid depression contributes to poor outcomes, and screening for depression in acute medical in-patients has been advocated. The Edinburgh Depression Scale (EDS) has been validated in a variety of general hospital patient groups, but not previously in older acute medical in-patients.
Methods: one hundred and eighteen patients aged 65 and older on acute medical wards were assessed using a standardised diagnostic interview (Present State Examination—Schedules for Clinical Assessment in Neuropsychiatry) to identify depression according to ICD-10 criteria. They subsequently completed the EDS. The performance characteristics at a range of thresholds were compared, and receiver operating characteristic curve analysis was performed.
Results: the optimal EDS cut-off for identifying ICD-10 depressive episode was 7/8, with a sensitivity of 88%, specificity of 77%, positive predictive value of 52% and negative predictive value of 96%. The area under the receiver operating characteristic curve was 0.91.
Conclusion: the EDS was shown to be a useful instrument for detecting clinical depression in older people on acute medical wards in this study. Its performance was equivalent to other validated screening instruments in this population. Our findings add further weight to using the EDS as a screening instrument for depression in multiple general hospital settings.