Delivering the Thinking Healthy Programme for perinatal depression through peers: an individually randomised controlled trial in India

Abstract

Background:The Thinking Healthy Programme (THP) is a psychological intervention recommended for the treatment of perinatal depression. We assessed the effectiveness and cost-effectiveness of THP delivered by peers (THPP) in Goa, India. Methods:In this single-blind, individually-randomised controlled trial, we recruited pregnant women aged ≥18 years attending antenatal clinics, who scored ≥10 on the Patient Health Questionnaire (PHQ-9). Participants were randomly allocated (1:1) to THPP plus enhanced usual care (EUC), or to EUC alone in randomly-sized blocks, stratified by area of residence. Allocation was concealed using sequentiallynumbered opaque envelopes. Primary outcomes were severity of depressive symptoms (PHQ-9 score) and remission (PHQ-9 score <5) 6 months post-birth assessed by researchers masked to the treatment allocation. Analyses were by intention to treat, adjusting for covariates defined a priori or that showed imbalance at baseline. The trial is registered with ClinicalTrials.gov (NCT02104232).Findings: 280 women were enrolled between 24th October 2014 and 22nd June 2016 (140 per group). At 6 months, 122 (87%) and 129 (92%) women in the THPP plus EUC, and EUC alone groups, respectively, contributed primary outcome data. There was evidence of higher prevalence of remission at 6 months in the THPP plus EUC versus EUC alone group (n=89 (73%) versus n=77 (60%), respectively; prevalence ratio (PR)=1·21; 95% confidence interval (CI) 1·01 to 1·45, p=0·04), but weaker evidence for lower symptom severity (mean 3·47, standard deviation (SD) 4·49 versus 4·48 (SD=5·11), respectively; standardised mean difference (SMD)=-0·18; 95% CI -0·43 to 0·07, p=0·16).Repeated measures analyses over the duration of the trial (measured at 3 and 6 months post-birth) showed beneficial effects on symptom severity (SMD -0·37, 95% CI -0·88 to -0·24, p=0·01), remission (PR=1·21, 95% CI 1·01 to 1·41, p=0·02), WHO-Disability Assessment Schedule (WHO-DAS) scores (SMD=-0·32, 95% CI -0·76 to -0·21, p=0·02) and Multidimensional Scale of Perceived Social Support (MSPSS) scores (SMD=0·51, 95% CI 0·43 to 1·20, p=0·02). . The incremental societal cost per unit improvement on PHQ-9 was -US$ 29·64 (95% CI-32·88 to -26·41) at 3 months, and -US$93·53 (95% CI -180·21 to -6·84) for the whole duration of the trial, with a 87% likelihood of being costsaving in the study setting. The mean cost of providing THPP was US$ 1·36 per beneficiary (95% CI 1·32 to 1·39). There was no evidence of differences in serious adverse events by group, except stigmatisation was more often reported in the EUC alone (9%) versus the THPP plus EUC group (4%; p=0·08).Interpretation: THPP had a moderate effect on symptom severity and remission from perinatal depression over the 6 month post-natal period. THPP is relatively cheap to deliver and pays for itself through reduced health care, time and productivity costs. Funding: NIMH

Perinatal mental health: a review of progress and challenges

“…Most trials have been conducted in a Western country (usually Australia, US or UK), but some high‐quality RCTs have also been carried out in low‐resource settings, documenting that CBT‐based interventions delivered by trained community mental health workers 96 or peers 97 can be effective and cost‐effective when compared to enhanced usual care only 98 , though this was not found in all settings 99 .…”

Psychological interventions for common mental disorders in women experiencing intimate partner violence in low-income and middle-income countries: a systematic review and meta-analysis

“…However, several were designed to be trauma-informed, by addressing cognitive processing of traumatic events, 21 habituating traumatic memories, 22 focusing on traumatic experiences, 29 or debriefing. 28 The remaining interventions had a more practical focus and were likely to benefit women disclosing intimate partner violence through problemsolving, 25,31 behavioural activation, 21,26,32,34 coping skills, 27,30 or a mixture of methods. 33 Given that anxiety is an important manifestation of trauma, more trauma-focused interventions might yield even greater gains than generic interventions in women experiencing intimate partner violence.…”

“…Included interventions were delivered by various staff: a public health professional and researcher, 28 psychology Masters-educated therapists, 29 community mental health workers, 21,30 general health workers, 31,33 other community workers, 27 and lay individuals. 22,25,26,32,34 Five studies [25][26][27]32,34 compared the control intervention (enhanced usual care [EUC]) with the intervention plus EUC, but the content of EUC, and other studies' control interventions, varied from face-to-face or telephone-delivered basic counselling 31,33 to diagnosis and the opportunity for WHO mental health gap action programme intervention guided treatment, 26,32,34 a home visit, 27 film screenings, 28 and waiting lists. 21,22,29,30 Numbers of sessions ranged from three 28 to 14, 33 delivered at clinics, 21,30 in community settings, 25,27,28 at home, 22,31,34 or a mixture of locations.…”

“…For example, studies measured postintervention outcomes at variable times (immediately, 28-30 3 months, 22,27,31,32 or 6 months 21,25 post-participation, and 3 months 33 or 6 months 26,34 postpartum) and varied in their use of validated intimate partner violence measures. Most studies asked a single question about intimate partner violence 32,33 or domestic violence 21,26,29,30,34 exposure, which might have included abuse by non-partners. Some studies 26,33,34 asked about intimate partner violence in the past 3 months only, perhaps underestimating intimate partner violence prevalence and one study 25 asked about domestic upheaval in the past 6 months, perhaps overestimating intimate partner violence prevalence.…”

“…Most studies asked a single question about intimate partner violence 32,33 or domestic violence 21,26,29,30,34 exposure, which might have included abuse by non-partners. Some studies 26,33,34 asked about intimate partner violence in the past 3 months only, perhaps underestimating intimate partner violence prevalence and one study 25 asked about domestic upheaval in the past 6 months, perhaps overestimating intimate partner violence prevalence. One study 32 measured intimate partner violence exposure at 3-month follow-up.…”