Valid assessment of the clinical features of depression by relatives appears to slip under the RADAR

OBJECTIVE: We report on the development of an observational measure designed for completion by relatives, the Recent Appearance of Depression Assessed by Relatives (RADAR), and consider its validity. METHOD: One hundred and one patients with a current major depressive episode had a relative or close friend complete the RADAR, while psychiatrists and research assistants collected extensive data. RESULTS: Correlated against both patient self-report and psychiatrist-rated depression severity measures, RADAR scores evidenced poor validity overall. Recent Appearance of Depression Assessed by Relatives scores also failed to differentiate depressive subtypes, in that scores for patients with melancholic depression were not significantly higher than for those with non-melancholic depression. CONCLUSIONS: Results are consistent with previous findings of poor agreement between clinicians and corroborative witnesses in assessing clinical depressive features, and argue against reliance on corroborative witness reports.