Effect of data-collection method on reporting of common mental disorder symptoms and intimate partner violence in Zimbabwe: a cluster-randomized trial.

BACKGROUND: Screening for sensitive and stigmatized conditions such as mental health or experience of violence is challenging. Audio computer-assisted self-interviewing (ACASI), administered by using a tablet and headphones, may be more sensitive for this purpose than paper-based self-administered questionnaires (SAQ) handed in to project staff. We conducted a methodological cluster-randomized trial in Zimbabwe to compare two methods of screening for common mental disorders (CMD) and intimate partner violence (IPV): ACASI versus SAQ. METHODS: Trial participants were health workers receiving occupational health checks at hospitals and primary health clinics. The unit of randomization was a working day. CMD was measured by using the Shona Symptom Questionnaire, anxiety by using the Generalised Anxiety Disorder-7 questionnaire, and IPV by using the World Health Organization screening questionnaire. The co-primary outcomes were CMD prevalence and the prevalence of any IPV, compared by arm at the cluster level, adjusting for gender and weekend. Secondary outcomes were the prevalence of anxiety and of physical, emotional, and severe physical and sexual IPV. RESULTS: Between 20 February and 10 June 2022, 1240 participants were enrolled in 71 clusters (workdays), with 77.0% female and 66.4% in clinical-facing roles. The cluster-level geometric mean prevalence of CMD was 19.4% when using ACASI and 14.1% when using SAQ [adjusted risk ratio (aRR) 1.37, 95% confidence interval (CI) 0.99, 1.89; P = .056]. ACASI yielded a higher prevalence of overall IPV than the SAQ (cluster-level geometric mean prevalence 40.6% compared with 22.4%, aRR 1.81, 95% CI 1.40, 2.35; P < .001), of emotional IPV (aRR 1.66, 95% CI 1.27, 2.17; P < .001), and of physical IPV (aRR 1.61, 95% CI 1.16, 2.25; P = .005). No differences were seen in the prevalence of severe physical or sexual IPV or anxiety across the trial arms. CONCLUSION: Screening for CMD and IPV by using a confidential ACASI method identifies more people who may benefit from care than screening by using SAQ handed in to clinic staff. This may be explained by under-reporting on the SAQ. ACASI is a promising screening method for sensitive issues in healthcare settings.