Detecting depression following a stroke

Abstract

The assessment of depression in patients following a stroke is complex with the consequence that there are few well validated tools to aid identification. The aim of this thesis was to examine a range of approaches to detecting depression in people who have suffered a stroke. A series of three studies was performed ( casenote review, feasibility study and main study) exploring the effectiveness of different strategies for diagnosing and screening for depression in hospitalised stroke patients. The casenote review included 159 patients; median age 74 years (IQR 68-79), 86 (54.1 %) were male. The Psychology Assistant assessed patients using a clinical interview, and found 78 (55%) patients were depressed. The MDT often identified depression in patients who were not depressed (sensitivity 96.6%; specificity 15.2%). It was also apparent that there was no systematic assessment or recording of depression following a stroke. The feasibility study piloted validation of the Geriatric Mental State [GMS-AGECAT] and the Montgomery Asberg Depression Rating Scale (MADRS) as diagnostic interviews when performed by the research nurse in comparison to the "gold standard" the Psychiatrist. In addition the utility of an observational screening tool, the Signs of Depression Scale [SODS] for mood problems in people who have had a stroke, including those with cognitive and communication problems, was explored. The feasibility study informed some minor changes to the main study, which was undertaken. Seventy-four patients were seen by both the research nurse and by the Psychiatrist and were therefore included in the main analysis. The median age was 70 (IQR, 59-78) and 40 (54%) were male. Nearly half (49%) had abnormal communication with only 10% having normal memory function. According to the Psychiatrist, 27 (36%) patients were depressed. According to the research nurse, the GMS AGECAT and the MADRS classified 32/74 (43%) and 38/71 (54%) patients respectively as depressed. Relative to the psychiatric interview, the GMS-AGECA T had a sensitivity of 66. 7% and a specificity of 70.2%. The MAD RS had a sensitivity of 80.8% and a specificity of 62.2%. Seventy-one patients had the SODS completed by a ward nurse, 55 by a therapist and 31 by a carer.