Self report questionnaire assessment of anxiety and depression amongst stroke patients in rehabilitation settings.
D Clin Psy thesis, University of Glasgow.
Full text available as:
Background and purpose: Depression and anxiety are common stroke sequelae and are detrimental
to outcomes if not detected and addressed. Some self-report measures of anxiety and depression
have been criticised for lack of specificity and face validity of item structure [Hospital Anxiety and
Depression Scale - HADS] or may not be fully validated for a stroke population [Geriatric Depression
Scale Short Form - GDS-SF]. A recently developed anxiety measure may be useful for this population
[Geriatric Anxiety Inventory - GAI]. The purpose of the study is to assess the clinical utility of these
measures for screening mood disorders in people over 45 years old undergoing stroke rehabilitation.
Methods: The HADS, GAI and GDS-SF were assessed against DSM-IV “gold standard” diagnoses from
the Mini International Neuropsychiatric Interview [MINI]. A sample of patients in rehabilitation
following stroke (n=34) was used. The age range was 46-92 (mean 73.12 years; SD=12.37). 21
participants were female and 13 were male (61.8% and 38.2% respectively). The study sample had
relatively intact cognitive function as assessed by referring clinicians and relatively high
communicative ability with Frenchay Aphasia Screening Test [FAST] scores in the range 18 to 30
(mean 26.38; SD=2.94).
Results: All measures were able to distinguish those with index disorders from those without. HADSA
displayed sensitivity 91%, specificity 70% at a cut-off of 8/21. HADS-D displayed sensitivity 82%,
specificity 83% at a cut-off of 8/21. GAI displayed sensitivity 91%, specificity 65% at a cut-off of 8/20.
GDS-SF displayed sensitivity 63%, specificity 87% at a cut-off of 8/15.
Conclusions: All conclusions are tentative as this is a small scale preliminary study. The HADS
performed best in screening for anxiety and depression, and is suitable for use with a cognitively
intact stroke population. The GAI also performed well and is suitable for use with this population.
The GDS-SF had poor sensitivity and so does not appear suitable for this population.
Actions (login required)