Search

Mental health conditions are common during the perinatal period, and associated with maternal, foetal, and neonatal morbidity and mortality. The prevalence of antenatal depressive symptoms in rural Ethiopia is estimated to be 12-29%. There is an established bidirectional relationship between mental ill-health and intimate partner violence (IPV), including during and after pregnancy. The lifetime prevalence of IPV exposure among women in rural Ethiopia is as high as 72%and may be even higher during the perinatal period. Although a growing literature demonstrates the efficacy of brief psychological interventions for the treatment of common mental disorders (CMDs), including clinically relevant depressive, anxiety, and post-traumatic stress disorder (PTSD) symptoms, in low and middle-income countries (LMICs), few measure or examine the impact on pregnant women, or women experiencing IPV, especially in rural, low-income country settings. Pregnancy is the commonest time for Ethiopian women to access healthcare, making antenatal care (ANC) an important opportunity for interventions that address the mental health impacts of IPV. The aims of this PhD were as follows. (1) To test the hypothesis that IPV exposure moderates the efficacy of psychological interventions for CMDs in LMICs, with IPV exposure associated with lower treatment effects(Study One). (2) To explore the perspectives of women and ANC staff on perinatal emotional difficulties, IPV, and features of an acceptable intervention in rural Ethiopia (Study Two). (3) To adapt a brief psychological intervention for the perinatal emotional difficulties of women experiencing IPV in rural Ethiopia (Study Three). (4) To evaluate the feasibility of the adapted intervention and a randomised, controlled study design (Studies Four and Five). In Study One, I conducted a systematic review and meta-analysis of randomised controlled trials of psychological interventions conducted in LMICs, which measured IPV exposure among women participants. Study Two comprised in-depth qualitative interviews with 16 pregnant women and 12 members of ANC staff in Sodo, Ethiopia, analysed using thematic analysis. Interviews explored participants' experiences of perinatal emotional difficulties and IPV, to inform the adaptation of a brief psychological intervention and its implementation strategies, for this context. In Study Three, I identified the psychological intervention model best suited to addressing the findings of Studies One and Two. I then followed the latest guidance to adapt problem-solving therapy (PST) for the perinatal emotional difficulties of women experiencing IPV in Sodo (PST-IPV), including devising a theory of change and corresponding 'dark logic model', to capture potential unintended adverse impacts. In Study Four, I developed the protocol for a randomised, controlled feasibility trial, comparing PST-IPV with standard PST (not adapted for women experiencing IPV), and enhanced usual care (EUC). In Study Five, I analysed the results for the 52 participants randomised to PST-IPV (n=25), standard PST (n=12), and EUC (n=15). Study One found, contrary to my hypothesis, that women disclosing IPV benefitted more from psychological interventions for CMDs than women who did not, in LMICs. This effect was particularly evident for anxiety symptoms (difference between standardised mean differences (dSMD)=0.31, CI:0.04, 0.57), with a similar but non-significant effect of IPV exposure on reductions in PTSD (dSMD=0.14, CI: -0.06, 0.33), depressive symptoms (dSMD=0.10, CI: -0.04, 0.25), and psychological distress (dSMD=0.07, CI:-0.05, 0.18.Study Two found that abusive treatment, emotional and bodily distress were widespread in Sodo, and that women and health workers were familiar with the relationship between them. Participants contextualised IPV as the primary form of abusive treatment women experienced, connected by multiple pathways to emotional and bodily distress. Patriarchal norms explained how the actions of neighbours, family, community leaders, law enforcement, and government agents in response to IPV often reinforced women's experiences of abuse. This created a sense of powerlessness, exacerbated by the tension between high cultural expectations of reciprocal generosity and severe deprivation. Participants advocated an ANC intervention with a problem-solving focus, addressing women's powerlessness, but they also identified intrinsic benefits of unburdening oneself and being listened to. Women and health workers made a range of recommendations, including prioritising confidentiality, building women's trust in ANC staff, and raising community awareness about the adverse effects of IPV. Study Three found that a model of brief PST developed in South Africa was best suited to the findings of Studies One and Two. I adapted the delivery format in line with women's prioritisation of confidentiality and brevity. I adapted the 'surface' of treatment delivery via translation, culturally congruent examples, locally relevant illustrations, conceptual simplification, removal of incongruent elements, developing an in-session flip-chart resource, and accommodation of non-biomedical explanatory models of distress. Adaptations for pregnant women in this context included integration into routine ANC appointments and adding relaxation techniques. Adaptations for the needs of women experiencing IPV comprised basic training on responding to IPV disclosures, based on international guidance, incorporating anonymised local women's IPV testimonies from Study Two, practice applying the PST model to IPV-related problems, clear guidance on how to manage IPV-related problems throughout the manual and training course, and safeguards for responding to IPV-related risks. The theory of change identified short, medium, and long-term outcomes required to achieve PST-IPV's desired impact, and the evidence and experience-informed actions required to achieve each step. It also articulated underlying assumptions and progress markers. Studies Four and Five found that attending four sessions of PST-IPV, and implementing arandomised, controlled study design, comparing PST-IPV with active and inactive arms, was acceptable to women and health workers, and feasible in Sodo. Evidence included a recruitment rate of 10 participants per week across two sites, 76% of participants randomised to PST-IPV completing four sessions, and 24% dropping out, usually due to moving away or postnatal confinement. Staggering recruitment in line with therapist availability to address the high recruitment rate, focusing therapist training on communication skills, automating randomisation decisions, training supervisors using a cascade model, and conducting immediate and longer-term outcome assessments were recommended adjustments. This PhD concluded that:(1) IPV exposure does not reduce the efficacy of psychological interventions for CMDs in LMICs. (2) Women and ANC staff in rural Ethiopia perceive multiple, interconnected pathways between perinatal emotional difficulties and IPV, for which brief psychological interventions are acceptable. (3) Brief PST can be adapted for pregnant women experiencing IPV by reviewing the context in depth, following recent adaptation guidance, and developing a theory of change with stakeholders. (4) PST-IPV and the randomised, controlled study design are acceptable to women and ANC staff, and feasible to implement. (5) Study Four's design requires adjustment for a future, fully-powered RCT. Limitations included the lack of availabledata from 29% of eligible papers in Study One, the inability to conduct member checking for Studies Two and Three, weakening of statistical power to detect between-arms differences in Study Four, due to a protocol deviation, and delays to qualitative interviews with Study Four stakeholders, caused by the coronavirus pandemic.