Added value of assessing medical students' reflective writings in communication skills training: a longitudinal study in four academic centres

Objectives This study describes the development and implementation of a model to assess students' communication skills highlighting the use of reflective writing. We aimed to evaluate the usefulness of the students' reflections in the assessment of communication skills. Design Third-year and fourth-year medical students enrolled in an elective course on clinical communication skills development were assessed using different assessment methods. Setting and participants The communication skills course was offered at four universities (three in Brazil and one in Portugal) and included 69 students. Outcome measures The students were assessed by a Multiple-Choice Questionnaire (MCQ), an objective structured clinical examination (OSCE) and reflective writing narratives. The Cronbach's alpha, dimensionality and the person's correlation were applied to evaluate the reliability of the assessment methods and their correlations. Reflective witting was assessed by applying the Reflection Evaluation for Enhanced Competencies Tool Rubric (Reflect Score (RS)) to measure reflections' depth, and the Thematic Score (TS) to map and grade reflections' themes. Results The Cronbach alpha for the MCQ, OSCE global score, TS and RS were, respectively, 0.697, 0.633, 0.784 and 0.850. The interobserver correlation for the TS and RS were, respectively, 0.907 and 0.816. The assessment of reflection using the TS was significantly correlated with the MCQ (r=0.412; p=0.019), OSCE (0.439; p=0.012) and RS (0.410; p=0.020). The RS did not correlate with the MCQ and OSCE. Conclusions Assessing reflection through mapping the themes and analysing the depth of reflective writing expands the assessment of communication skills. While the assessment of reflective themes is related to the cognitive and behavioural domains of learning, the reflective depth seems to be a specific competence, not correlated with other assessment methods-possibly a metacognitive domain.