Application of Research Design and Methods for Optimizing Prevention Science: Mechanisms of Change in a Community-Based Family Intervention in Kenya

This abstract was presented at the 2018 Society for Prevention Research Annual Meeting which was held May 29 – June 1, 2018 in Washington, DC, US.

Eve S. Puffer Duke University

Ali Giusto Duke University; Elsa A. Friis Duke University; Bronwyn N. Kaiser Duke University; David Ayuku Moi University School of Medicine

Background: Family-based interventions hold promise for mitigating effects of contextual stressors, including poverty, on parental violence and children’s mental health. Current evidence supports efficacy of family-based interventions, but less is known about mechanisms of change. 

Aim: To explore mechanisms underlying effects of a family intervention in Kenya. 

Methods: A pilot pre-post study was conducted (N= 14 families). The intervention, delivered by lay counselors, combines evidence-based practices: solution-focused family therapy, behavioral skills training, and cognitive-behavioral strategies. These are streamlined into a flexible components-based intervention. After pre-post comparisons showed family functioning and mental health improvements., mechanisms were explored using a mixed methods approach. Analyses were conducted in two phases. First, we used structured mapping of clinical change per family and across cases through analysis of session transcripts, surveys, and qualitative interviews. This identified family interaction processesdriving improvements in family functioning and mental health. Second, we analyzed transcripts and ratings of counselor fidelity and competency to identify elements of intervention content and delivery that seemed to drive these change processes. 

Results: Pre-intervention, common triggers for dysfunctional interaction patterns included lack of resources and alcohol consumption, both often a cause or result of caregiver mental health symptoms. Together, these contributed to marital and parent-child relationship problems characterized by conflict/violence and lack of communication and trust, leading to child mental health symptoms. During and post-treatment, processes of change highlighted intersections between family interactions, mental health, and poverty-related stress. Three related pathways included: (1) Marital communication improvements led to decreased violence, increased trust, and better financial problem-solving; these improved caregiver mental health; (2) Decreased marital violence facilitated improved parent-child relationships and decreased child maltreatment, both improving child mental health; and (3) Improved parent-child communication (often related to material needs), paired with clear expectations, decreased maltreatment and improved child behavior. Drivers of change related to content included communication skills training, facilitation of communication during sessions focused on solution-finding and empathy, homework focused on positive time, parent psychoeducation on positive attention and expectations, and behavioral coping skills training. Drivers related to counselor competencies included facilitating participation of all members, building trusting relationships, and integrating religious or local examples to emphasize hope and change. 

Discussion: The identified processes of change help generate hypotheses about which components and counselor delivery strategies are most essential for achieving outcomes. Findings are particularly relevant for low-resource settings given the need to consider poverty-related stress in presenting problems, change processes, and implementation.