Evaluation of the postgraduate family medicine decentralised training programme at the university of Witwatersrand, South Africa, using the logic model

Abstract

Postgraduate family medicine decentralised training programmes were implemented in South Africa about 15 years ago, but the University of the Witwatersrand’s programme has not been comprehensively evaluated. This study evaluated the programme using a complex programme evaluation logic model based on linearity theory. This theory assumes ideal inputs and processes produce good programmatic outputs and outcomes. Resources and support were evaluated as inputs, postgraduate supervision and workplace-based learning as processes, supervisory feedback as outputs and workplace-based assessments as outcomes. A parallel convergent mixed-methods instrumental case study was conducted with purposively-sampled family physicians (n=11) and trainees (n=11) from five decentralised training sites. Semi-structured interviews were audio recorded, transcribed verbatim, and analysed inductively using MAXQDA 2020 software. Descriptive statistical analysis was conducted on components of registrars’ learning portfolios (scores, supervisory feedback, and skills competence) and examination results using Stata 14.2 software. An integrative analysis involving transforming the quantitative results to qualitative findings and drawing meta-inferences was conducted. The integrated findings were used to modify the initial logic model and identify key recommendations to optimise the programme. The integrative analysis identified the need for more material and human resources, university and district management support, and standardised resources, supervision, and learning practices. Supervisors’ knowledge, skills, and behaviours varied across sites and their feedback was insufficient regarding soft skills like clinical reasoning and patient negotiation. Workplace-based assessments did not meet the required standards across training years and districts. Interpersonal interactions with patients, peers, supervisors and other professionals, engagement in district activities, promoted learning. Registrars’ professionalism and self-learning need improvement. The key recommendations include more explicit national guidelines, sufficient support from the provincial department, university, and district management, well maintained infrastructure, sufficient skilled supervisors, more professional development training for supervisors, protected time for registrar learning, and better use of self-learning and reflection.

Emulating successful contextual adaptations while addressing challenges across sites contributes to thriving decentralised training programmes in district health systems. An improved understanding of the concepts and their interrelationships in training programmes could be translated to similar decentralised training platforms across medical disciplines of sub-Saharan Africa or low-middle income countries