RESUMO
BACKGROUND: There are gaps in our understanding of how non-specialists, such as lay health workers, can achieve core competencies to deliver psychosocial interventions in low- and middle-income countries. METHODS: We conducted a 12-month mixed-methods study alongside the Rehabilitation Intervention for people with Schizophrenia in Ethiopia (RISE) pilot study. We rated a total of 30 role-plays and 55 clinical encounters of ten community-based rehabilitation (CBR) lay workers using an Ethiopian adaptation of the ENhancing Assessment of Common Therapeutic factors (ENACT) structured observational rating scale. To explore factors influencing competence, six focus group discussions and four in-depth interviews were conducted with 11 CBR workers and two supervisors at three time-points. We conducted a thematic analysis and triangulated the qualitative and quantitative data. RESULTS: There were improvements in CBR worker competence throughout the training and 12-month pilot study. Therapeutic alliance competencies (e.g., empathy) saw the earliest improvements. Competencies in personal factors (e.g., substance use) and external factors (e.g., assessing social networks) were initially rated lower, but scores improved during the pilot. Problem-solving and giving advice competencies saw the least improvements overall. Multimodal training, including role-plays, field work and group discussions, contributed to early development of competence. Initial stigma towards CBR participants was reduced through contact. Over time CBR workers occupied dual roles of expert and close friend for the people with schizophrenia in the programme. Competence was sustained through peer supervision, which also supported wellbeing. More intensive specialist supervision was needed. CONCLUSION: It is possible to equip lay health workers with the core competencies to deliver a psychosocial intervention for people with schizophrenia in a low-income setting. A prolonged period of work experience is needed to develop advanced skills such as problem-solving. A structured intervention with clear protocols, combined with peer supervision to support wellbeing, is recommended for good quality intervention delivery. Repeated ENACT assessments can feasibly and successfully be used to identify areas needing improvement and to guide on-going training and supervision.