Using learning analytics in clinical competency committees: Increasing the impact of competency-based medical...

Graduate medical education (GME) and Clinical Competency Committees (CCC) have been evolving to monitor trainee progression using competency-based medical education principles and outcomes, though evidence suggests CCCs fall short of this goal. Challenges include that evaluation data are often incomplete, insufficient, poorly aligned with performance, conflicting or of unknown quality, and CCCs struggle to organize, analyze, visualize, and integrate data elements across sources, collection methods, contexts, and time-periods, which makes advancement decisions difficult. Learning analytics have significant potential to improve competence committee decision making, yet their use is not yet commonplace. Learning analytics (LA) is the interpretation of multiple data sources gathered on trainees to assess academic progress, predict future performance, and identify potential issues to be addressed with feedback and individualized learning plans. What distinguishes LA from other educational approaches is systematic data collection and advanced digital interpretation and visualization to inform educational systems. These data are necessary to: 1) fully understand educational contexts and guide improvements; 2) advance proficiency among stakeholders to make ethical and accurate summative decisions; and 3) clearly communicate methods, findings, and actionable recommendations for a range of educational stakeholders. The ACGME released the third edition CCC Guidebook for Programs in 2020 and the 2021 Milestones 2.0 supplement of the Journal of Graduate Medical Education (JGME Supplement) presented important papers that describe evaluation and implementation features of effective CCCs. Principles of LA underpin national GME outcomes data and training across specialties; however, little guidance currently exists on how GME programs can use LA to improve the CCC process. Here we outline recommendations for implementing learning analytics for supporting decision making on trainee progress in two areas: 1) Data Quality and Decision Making, and 2) Educator Development.