Relationships between general practice residents’ level of reflection and their competencies development: results from a pilot cross-sectional study

Abstract

Reflective practice is a cornerstone for developing professional competence. Despite its importance, the relationship between levels of reflection and competence remains underexplored in medical education. This association was described in trainee teachers. Along with initiating a cohort study to follow the development of reflection and competencies in general practice (GP) residents, we conducted this pilot study to evaluate the data collection and methodological approach in terms of feasibility and validity of the assessment. The objective of the study was to assess the relationship between the level of reflection in GP residents' reflective writings and their competencies levels at the end of their first six-month internship. In April–May 2022, GP residents’ competencies levels were assessed by their clinical supervisors in three French medical schools. Five medical teachers assessed residents’ reflective narrative using the REFLECT rubric. We explored associations between competencies and reflection levels using Fisher’s exact test and logistic regression model. We also analyzed validity evidence following Messick’s framework. No association was found between reflection and competencies levels except for the “Primary care, emergency” competence, with more students assessed at the competent level in the lower levels of reflection than in the higher levels of reflection. This effect pattern was confirmed by the logistic regression model, for the PCE competence ( n = 11, 34.4% vs n = 1, 5.6%; OR = 0.11 [CI95 0.01 – 0.96]) and for all competencies, notably EPICH ( n = 14, 43.8% vs n = 2, 11.1%; OR = 0.16 [CI95 0.03 – 0.82]). Validity threats were primarily linked to response processes, notably very low inter-rater agreement for reflection assessment, heterogeneity in workplace-based assessment practices, and aggregation rules that may have compressed scoring variability. This pilot study did not demonstrate robust association between reflection and competencies levels at a single time point in the early stages of residency. Findings highlight key measurement constraints and feasibility issues when using single-method, cross-sectional quantitative approaches to assess complex constructs. The study emphasises the importance of improving assessment processes and validity evidence before drawing inferences about correlations and supports the use of longitudinal and pluralistic methodological approaches to better capture the dynamic relationship between reflection and competence development.