REVISION AND PSYCHOMETRIC PROPERTIES OF THE SIDDIQUI-SHAH DEPRESSION SCALE (SSDS)

Abstract

The Siddiqui Shah Depression Scale (SSDS) is an indigenous screening tool for depression developed to address gaps in treatment provision in Pakistan, that has been widely utilized locally in both clinical and research settings. However, changes in cultural idioms of distress, linguistic expression, and advances in psychometric analyses since development has rendered it in need of revision and revalidation. The current study aims to revise the SSDS in line with best practices to enhance the effectiveness of the screening instrument. A total of 274 adults were recruited for the study, with 174 clinical cases and 100 non-case controls from outpatient psychiatry departments in major hospitals across Rawalpindi and Islamabad. The research employed a cross-sectional correlational design with three phases. In the first phase, the original SSDS was revised for construct-irrelevant source of variance in items, instructions and scoring methods. The revised SSDS (SSDS-R) consisted of 34 items with item and instruction changes. In the second phase, the main study instruments including SSDS-R and Urdu WHO Quality of Life-Brief (WHOQOL-BREF) were pilot tested for applicability to the study sample and internal consistency, all of which demonstrated adequate reliability. The third phase of the study focused on establishing relevant psychometric properties of the SSDS-R. Results indicated SSDS-R had strong internal consistency at α=.90 for the clinical sample, with a split half value of α=.87. Weak negative correlation existed with the domains of WHOQOL-BREF, with psychological domain r=-.41 (p<0.01), physical domain r=-.39 (p<0.01), social domain r=-.34 (p<0.01), and environmental domain r=-.29 (p<0.01). Factor structure was explored with Exploratory xv Factor Analysis (EFA), utilizing Principal Component Analysis. A unidimensional single-factor structure was retained with consensus of Kaiser Criterion, Scree Plot, Velicer’s MAP, and Parallel Analysis, accounting for 24.7% of the variance. Results of the ROC curve analysis suggested an optimal cutoff at 41 utilizing Youden’s Index with sensitivity of .74 and specificity of .82 (PPV=.87, NPV=.65). The data therefore suggests that the SSDS-R is a reliable and valid indigenous screening tool with adequate diagnostic utility for bridging the treatment gaps of healthcare systems in Pakistan.