Comparison of assessment measures for post-stroke depression

Depression is commonplace following stroke and is associated with increased morbidity and mortality. The objective of this study is to examine the sensitivity and specificity of four standardized measures for assessing depression in a stroke population. A total of 67 adults admitted to an inpatient rehabilitation unit with acute stroke were assessed with four standardized depression measures and by structured clinical interview as part of a cross-sectional study assessing cognitive and emotional functioning. Using DSM-IV criteria, major depression was present in 15% of the sample and minor depression in an additional 28%. The four depression measures yielded rates of depression ranging from 14% to 46% when their recommended clinical threshold scores were applied. The Center for Epidemiologic Studies Depression Scale (CES-D) had the greatest positive predictive value (PPV) at 28%. Cutoff scores were adjusted using receiver operating curve (ROC) analyses, and PPV ranged from 28% to 34% when using lower cutoff scores, representing an improvement for all measures. Common assessment measures yield significantly different classification rates for depression in persons with acute stroke. Traditional threshold scores for commonly used objective depression measures may not be optimally sensitive for detection of depression in stroke populations. Based on the findings of this study we recommend clinicians use the Geriatric Depression Scale-Short Form using a cutoff score of 3 or greater. Clinical and research implications are offered.