Implementing psychosocial support models in contexts of extreme adversity: lessons from a process evaluation in Colombia.

Introduction: High quality investments during early childhood allow children to achieve their full potential by setting developmental foundations. However, challenges in the scale-up of evidence-based interventions make across-the-board implementation a non-trivial matter. Moreover, extreme contextual conditions -such as community violence, forced displacement, and poverty- impose a double threat. First, by directly affecting early childhood development (ECD), forced displacement and exposure to violence during early childhood, coupled with deficits in nurturing relationships, can trigger toxic stress, affecting children's mental health and social and emotional learning. Second, contexts of extreme adversity exacerbate common implementation pitfalls in the scale-up of interventions. Recognizing and documenting "what it takes" to successfully implement "what works" can contribute to the expansion and effectiveness of evidence-based programs that promote ECD in these settings. Methods: This article presents the results of the process evaluation of SA during the 2018-2019 implementation in Tumaco, a violence ridden municipality in the south-west border of Colombia, South America. In this phase, the program reached 714 families, 82% direct victims of violence and 57% were internally displaced. The process evaluation combined qualitative and quantitative methodological approaches to produce evidence of factors that promoted implementation quality. Results: Findings identified salient components of the program that promoted the program's acceptability, adoption, appropriateness, fidelity and sustainability: a rigorous cultural adaptation; well-structured team selection and training methodologies; and a team support and supervision protocol to provide continuous capacity building and prevent burn-out and other occupational hazards common among professionals in mental health and psychosocial support interventions. The statistical analysis using monitoring data identified key predictors of the dosage delivered (a measure of fidelity). Evidence suggests that initial attendance to the program and observable characteristics -such as educational attainment, violence victimization and employment status-predict a successful compliance (in terms of dosage to benefit from the program). Discussion: This study provides evidence for the development of structural, organizational, and procedural processes for the adoption, appropriate adaptation, and high-fidelity delivery of psychosocial support models delivered in territories affected by extreme adversity.