Assessment of perinatal anxiety: diagnostic accuracy of five measures

Citation

Ayers S, Coates R, Sinesi A, Cheyne H, Maxwell M, Best C, McNicol S, Williams LR, Uddin N, Hutton U, Howard G, Shakespeare J, Walker JJ, Alderdice F & Jomeen J (2024) Assessment of perinatal anxiety: diagnostic accuracy of five measures. The British Journal of Psychiatry. https://doi.org/10.1192/bjp.2023.174

Abstract
Background Anxiety in pregnancy and after giving birth (the perinatal period) is highly prevalent but under-recognised. Robust methods of assessing perinatal anxiety are essential for services to identify and treat women appropriately. Aims To determine which assessment measures are most psychometrically robust and effective at identifying women with perinatal anxiety (primary objective) and depression (secondary objective). Method We conducted a prospective longitudinal cohort study of 2243 women who completed five measures of anxiety and depression (Generalized Anxiety Disorder scale (GAD) two- and seven-item versions; Whooley questions; Clinical Outcomes in Routine Evaluation (CORE-10); and Stirling Antenatal Anxiety Scale (SAAS)) during pregnancy (15 weeks, 22 weeks and 31 weeks) and after birth (6 weeks). To assess diagnostic accuracy a sample of 403 participants completed modules of the Mini-International Neuropsychiatric Interview (MINI). Results The best diagnostic accuracy for anxiety was shown by the CORE-10 and SAAS. The best diagnostic accuracy for depression was shown by the CORE-10, SAAS and Whooley questions, although the SAAS had lower specificity. The same cut-off scores for each measure were optimal for identifying anxiety or depression (SAAS ≥9; CORE-10 ≥9; Whooley ≥1). All measures were psychometrically robust, with good internal consistency, convergent validity and unidimensional factor structure. Conclusions This study identified robust and effective methods of assessing perinatal anxiety and depression. We recommend using the CORE-10 or SAAS to assess perinatal anxiety and the CORE-10 or Whooley questions to assess depression. The GAD-2 and GAD-7 did not perform as well as other measures and optimal cut-offs were lower than currently recommended.

Keywords
Anxiety or fear-related disorders; perinatal psychiatry; rating scales; diagnostic accuracy; depressive disorders

Notes
Additional authors: the MAP study Team

Journal
The British Journal of Psychiatry

People

Dr Catherine Best

Lecturer Statistician, Institute for Social Marketing

Professor Helen Cheyne

Personal Chair, NMAHP

Professor Margaret Maxwell

Director of NMAHP Research Unit, NMAHP

Miss Stacey McNicol

Research Fellow, NMAHP