A comparative psychometric evaluation of the Edinburgh Postnatal Depression Scale and the Montgomery-Åsberg Depression Rating Scale-Self Report as measures of depression during pregnancy

Le résumé fourni par la source

Introduction The Edinburgh Postnatal Depression Scale (EPDS) is the most widely used instrument in perinatal clinical practice and research based on a belief that somatic symptoms during the perinatal period impact on depression measurement validity and reliability. The Montgomery Åsberg Depression Rating Scale – Self report (MADRS-S) may have better psychometric performance as a screening and outcome instrument for depression. Methods Self-referred pregnant women were screened (n=650), diagnostically assessed (n=390), treated with internet-based cognitive behavior therapy for depression (n=270) and followed-up diagnostically (n=169). EPDS and MADRS-S were assessed at screening and after treatment. Item level data from all patients were used in a series of psychometric tests including confirmatory factor analysis (CFA) and item response-theory tests (IRT). Results Both scales showed adequate reliability. Using depression as a unidimensional latent trait in CFA, MADRS-S fitted data better than EPDS. IRT revealed that the somatic items in MADRS-S contributed less than average rather than inflating the totals. For more depressed respondents, the EPDS scores’ validity decreased compared to MADRS-S. Conclusions MADRS-S and EPDS are reliable tools for assessing depressive symptoms in pregnant patients, but MADRS-S outperforms EPDS on several key psychometric indicators. Contrary to common belief, somatic items seem not to inflate scores, and MADRS-S retains better precision at higher symptom levels. These findings support the use of standard depression measures, such as the MADRS-S, in pregnant populations. Doing so may improve detection and treatment by integrating antenatal depression into general psychiatric care and research rather than isolating it within perinatal-specific frameworks.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.