Open Access Open Badges Research article

Quality of written narrative feedback and reflection in a modified mini-clinical evaluation exercise: an observational study

Elisabeth AM Pelgrim1*, Anneke WM Kramer1, Henk GA Mokkink1 and Cees PM Van der Vleuten2345

Author Affiliations

1 Department of Primary Care and Community Care, Radboud University Nijmegen Medical Centre, Postbus 9101 Huispostnummer 117, Nijmegen, HB, 6500, The Netherlands

2 Department of Educational Development and Research, Faculty of Health, Medicine, and Life Sciences, Maastricht University, Maastricht, The Netherlands

3 Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands

4 King Saud University, Riyadh, Saudi Arabia

5 University of Copenhagen, Copenhagen, Denmark

For all author emails, please log on.

BMC Medical Education 2012, 12:97  doi:10.1186/1472-6920-12-97

Published: 18 October 2012



Research has shown that narrative feedback, (self) reflections and a plan to undertake and evaluate improvements are key factors for effective feedback on clinical performance. We investigated the quantity of narrative comments comprising feedback (by trainers), self-reflections (by trainees) and action plans (by trainer and trainee) entered on a mini-CEX form that was modified for use in general practice training and to encourage trainers and trainees to provide narrative comments. In view of the importance of specificity as an indicator of feedback quality, we additionally examined the specificity of the comments.


We collected and analysed modified mini-CEX forms completed by GP trainers and trainees. Since each trainee has the same trainer for the duration of one year, we used trainer-trainee pairs as the unit of analysis. We determined for all forms the frequency of the different types of narrative comments and rated their specificity on a three-point scale: specific, moderately specific, not specific. Specificity was compared between trainee-trainer pairs.


We collected 485 completed modified mini-CEX forms from 54 trainees (mean of 8.8 forms per trainee; range 1–23; SD 5.6). Trainer feedback was more frequently provided than trainee self-reflections, and action plans were very rare. The comments were generally specific, but showed large differences between trainee-trainer pairs.


The frequency of self-reflection and action plans varied, all comments were generally specific and there were substantial and consistent differences between trainee-trainer pairs in the specificity of comments. We therefore conclude that feedback is not so much determined by the instrument as by the users. Interventions to improve the educational effects of the feedback procedure should therefore focus more on the users than on the instruments.