Evaluating Academic Mentorship Programs in Low- and Middle-Income Country Institutions: Proposed Framework and Metrics

Significant investments have been made to foster mentorship across a range of scientific disciplines. Many examples can be found in the medical and public health literature, including university programs,1–3 mentored fellowships,4,5 and multi-institutional collaborative networks.6–8 Underpinning these efforts is the understanding that contextual factors—including the institutional environment—can play a determining role in mentorship success. Despite this wealth of experience, however, there remains limited data to support specific mentoring practices. In their review of 39 studies, for example, Sambunjak et al.9 found that the evidence supporting the effectiveness of mentorship was not strong. They attributed this result to the poor quality of assessment data, including a predominance of cross-sectional studies, reliance on self-report surveys, lack of comparison groups, insufficient detail about the nature of mentoring relationships, and a focus on high-income country (HIC) settings. Other systematic reviews have reached similar conclusions. Across 13 mentoring programs for underrepresented minorities, Beech et al.10 described a relative paucity of outcome-driven evaluations. In a review of 18 studies, Kashiwagi et al.11 reported a larger number of evaluation metrics (e.g., mentee surveys, meeting attendance, retention rates, and academic productivity), but few long-term results. Formal mentoring in low- and middle-income country (LMIC) institutions has grown and these programs often face opportunities and challenges very different from their HIC counterparts.12,13 Unfortunately, the evaluation of many programs remains limited and this represents an important gap. Evaluation activities can help establish best practices across a variety of different settings. Performed on an ongoing basis, they can also be used to systematically track progress, identify programmatic gaps, and support efforts to improve the quality of mentorship.14 Program evaluations can also be used to justify past investments and leverage additional resources. At this time, few—if any—consensus metrics have been recommended to assess mentorship at an institutional level, even in HIC settings such as the United States.15 In this report, we describe a framework for the monitoring of mentorship programs, with a particular emphasis on LMIC institutions.