ABSTRACT
BACKGROUND: Uncertainty remains regarding the validity of screening tools to detect common mental disorders (CMDs) during perinatal periods. This umbrella review aims to provide an up-to-date summary of psychometric properties of tools for the identification of perinatal CMDs. METHODS: Reviews were identified via Ovid MEDLINE, PsychINFO, EMBASE, Global Health and Cochrane Database of Systematic Reviews electronic databases with no date or language restriction. Pooled sensitivity and specificity estimates and ranges were extracted and summarised using forest plots. Quality assessment was conducted using Measurement Tool to Assess Systematic Reviews (AMSTAR-2). RESULTS: Of 7,891 papers identified, 31 reviews met inclusion criteria. 76 screening tools were identified; most frequently validated were Edinburgh Postnatal Depression Scale (EPDS) (n = 28 reviews), Beck's Depression Inventory (BDI) (n = 13 reviews) and Patient Health Questionnaire (PHQ) (n = 12 reviews). Forest plots demonstrated a pattern of decreasing sensitivity and increasing specificity with increasing cut-off scores. Sub-group analysis of data extracted from low quality reviews demonstrated wider 95% CIs and overall lower specificity. Validity also varied according to ethnicity, socio-economic background and age. LIMITATIONS: Despite a low Covered Corrected Area (CCA) score the primary studies included within reviews overlapped; therefore we were unable perform meta-analysis. CONCLUSIONS: The evidence suggests that the EPDS, PHQ and BDI are useful across a range of diverse settings but the context of tool application is a key factor determining validity. This review highlights that utilizing screening tools in clinical practice is complex and requires careful consideration of the population, context, and health system it will be used in.
