Identifying perinatal depression with case-finding instruments: a mixed methods study (BaBY PaNDA - Born and Bred in Yorkshire PeriNatal Depression Diagnostic Accuracy)

Background Perinatal depression is well recognised as a mental health condition but Objectives To determine the diagnostic accuracy, acceptability and cost-effectiveness of the Whooley questions and the Edinburgh Postnatal Depression Scale (EPDS) to identify perinatal depression. Design A prospective diagnostic accuracy cohort study, with concurrent qualitative and economic evaluations. Setting Maternity services in England. Participants A total of 391 pregnant women. Main outcome measures Women completed the Whooley questions, EPDS and a diagnostic reference standard (Clinical Interview Schedule – Revised) during pregnancy (20 weeks) and postnatally (3–4 months). Qualitative interviews were conducted with health professionals (HPs) and a subsample of women. Results Diagnostic accuracy results: depression prevalence rates were 10.3% during pregnancy and 10.5% postnatally. The Whooley questions and EPDS (cut-off point of ≥ 10) performed reasonably well, with comparable sensitivity [pregnancy: Whooley questions 85.0%, 95% confidence interval (CI) 70.2% to 94.3%; EPDS 82.5%, 95% CI 67.2% to 92.7%; postnatally: Whooley questions 85.7%, 95% CI 69.7% to 95.2%; EPDS 82.9%, 95% CI 66.4% to 93.4%] and specificity (pregnancy: Whooley questions 83.7%, 95% CI 79.4% to 87.4%; EPDS 86.6%, 95% CI 82.5% to 90.0%; postnatally: Whooley questions 80.6%, 95% CI 75.7% to 84.9%; EPDS 87.6%, 95% CI 83.3% to 91.1%). Diagnostic accuracy of the EPDS (cut-off point of ≥ 13) was poor at both time points (pregnancy: sensitivity 45%, 95% CI 29.3% to 61.5%, and specificity 95.7%, 95% CI 93.0% to 97.6%; postnatally: sensitivity 62.9%, 95% CI 44.9% to 78.5%, and specificity 95.7%, 95% CI 92.7% to 97.7%). Qualitative evaluation: women and HPs were supportive of screening/case-finding for perinatal depression. The EPDS was preferred to the Whooley questions by women and HPs, mainly because of its ‘softer’ wording. Whooley question 1 was thought to be less acceptable, largely because of the terms ‘depressed’ and ‘hopeless’, leading to women not revealing their depressive symptoms. HPs identified a ‘patient-centred’ environment that focused on the mother and baby to promote discussion about mental health. Cost-effectiveness results: screening/case-finding using the Whooley questions or the EPDS alone was not the most cost-effective strategy. A two-stage strategy, ‘Whooley questions followed by the Patient Health Questionnaire’ (a measure assessing depression symptomatology), was the most cost-effective strategy in the range between £20,000 and £30,000 per quality-adjusted life-year in both the prenatal and postnatal decision models. Limitations Perinatal depression diagnosis was not cross-referenced with women’s medical records so the proportion of new cases identified is unknown. The clinical effectiveness and cost-effectiveness of screening/case-finding strategies was not assessed as part of a randomised controlled trial. Conclusions The Whooley questions and EPDS had acceptable sensitivity and specificity, but their use in practice might be limited by low predictive value and variation in their acceptability. A two-stage strategy was more cost-effective than single-stage strategies. Neither case-finding instrument met National Screening Committee criteria. Future work The yield of screening/case-finding should be established with reference to health-care records. The clinical effectiveness and cost-effectiveness of screening/case-finding for perinatal depression needs to be tested in a randomised controlled trial. Funding The National Institute for Health Research Health Services and Delivery Research programme.