Feedback is a vital aspect of supervision in clinical settings.^1,2 Feedback is defined as ‘a way the trainee comes to know about the gaps in their current level of knowledge and the desired goal’.³ For maximum effectiveness, feedback should be provided timeously, in small, manageable quantities, relate to observable behaviours⁴ and should be tailored to the trainee’s learning needs and non-judgemental.^5,6 Feedback must also be clearly articulated based on trainee’s observed performance,⁷ promote self-assessment,⁸ be context-specific, enable interactions between the trainee and supervisor³ and focus on adequate performance and areas for improvement.^9,10 Supervisors should assist trainees in developing written action plans specifying the supervisor’s role in supporting trainees to achieve the goals outlined in the learning plans.^9,10 More recent definitions regard feedback as a coaching activity where supervisors facilitate trainees’ self-reflection to improve performance.^10,11 However, the acceptability of feedback by trainees still depends on who provides it, whether it is constructive, the trainee’s assessment literacy and the nature of supervisor-trainee relationships.^5,11

When the feedback in the 36 learning plans was analysed, most (61.2%) did not refer to registrars’ clinical knowledge or skills, and 86.1% did not offer an improvement strategy. The ‘extent of supervisors’ feedback’ was mainly rated as ‘poor’ (47.2%) or ‘average’ (47.2%); only a few (5.6%) were ‘good’. In one-third of the forms, the final feedback was provided by non-FP supervisors from other clinical departments (Table 1). The data were not uniformly distributed, so the global score median of 8.0 (7.7–9.0) out of 10.0 in the learning plans was calculated.

In 17 of the 19 LPs, the mini-CEX forms were completed by more than one FP from the same district, while the same FP completed the remaining two. For the ‘extent of supervisors’ feedback’, 63.2% were classified as ‘poor’ and only 14.0% as ‘good’ (Table 2). The median of the total scores in the mini-CEX forms was 6.8 (range: 6.1–7.3). Most of the mini-CEX competency scores given by FP supervisors ranged between 5–7 or 8–10, with very few scores of 1–4 (Table 3).

Further analysis showed a statistically significant association of the ‘some feedback’ and ‘no feedback’ categories with all Mini-CEX competencies (p < 0.001 to p = 0.014), the rating of ‘extent of supervisors’ feedback’ (p < 0.001) and the ‘signed by registrar and supervisor’ (p < 0.001) categories with the training district (Table 4). However, there was no association with the year of training. There was no significant association of feedback categories in the learning plans with either the training district or year of training.

When the strength of association of D1–D5 with the feedback categories was tested further, D3 feedback rated higher when compared with other districts in all mini-CEX competencies with a statistical significance (p < 0.001 to p = 0.003). The ‘extent of supervisors’ feedback was also rated higher in D2 and D3 than D1, D4 and D5, which was significantly different (p < 0.001 to p = 0.002). D1 had significantly more signed and completed forms than D5 (p < 0.001).

Comparing the global scores in learning plans and total scores in mini-CEX forms with the year of training revealed significant differences in the median values of the global score (p = 0.012) and the total score (p = 0.049). On further analysis with two-sample Wilcoxon rank-sum tests, the differences were between Y1 and Y3 in both global score (p = 0.003) and the total score (p = 0.016). There were no significant differences between Y1 and Y2, or Y2 and Y3, across both forms. There were no significant differences in the median global and total scores across D1–D5.

The FP supervisors had provided the feedback in 22/36 learning plans, non-FP supervisors in 12/36 and 2/36 had no feedback. Non-FP supervisors provided more extended feedback (average of −25.8 words) than FP supervisors (average of −13.7 words). The feedback focused mainly on ‘highlighting strengths’ or was ‘non-specific’, with less emphasis on ‘specific areas for improvement’ and ‘providing action plans’ (Figure 1). The FP feedback focused only on registrars’ ‘work ethic’ and ‘clinical skills’. In contrast, non-FP feedback was more comprehensive because it referred to various areas, including ‘professional behaviour’ and ‘teamwork’ (Figure 2).

The FPs mini-CEX feedback focused on history-taking, comprehensive management and organisational efficiency. There was less emphasis on negotiation skills and comprehensive assessment (Figure 3). The average feedback by FPs in the mini-CEX forms (16.3 words) was longer than the learning plans (13.7 words). There was more emphasis on specifying ‘areas of improvement’ (per the overall quality evaluation) compared with the learning plans (Figure 1).

This is one of the first sub-Saharan studies evaluating feedback in registrars’ learning plans and mini-CEX forms as evidence of educational and clinical supervision. The general quantity and quality of the feedback across both forms were inadequate, which has implications for the registrars’ ability to engage in self-directed learning.

The feedback in the learning plans lacked elements of effectiveness such as context specificity, being tailored to the trainee’s needs and documented action plans. Other deficiencies were failure to focus on areas needing improvement, such as gaps in clinical knowledge, skills and professional behaviour and not specifying how the supervisor would assist the registrar in addressing these gaps. These findings were similar to previous studies where supervisors struggled to point out registrars’ weaknesses or offer suggestions for improvement or action plans.²⁹ Given feedback’s vital role in promoting self-directed learning,¹⁸ the lack of focus on registrars’ learning needs deprives them of opportunities to reflect on how they could improve their performance. As found previously, the written narrative feedback also minimally reflected registrar responses to feedback.^9,18

Several studies evaluating mini-CEX forms in HIC^23,39 found that feedback focused more on adequately and less on inadequately performed areas and did not always include documented action plans. The findings of this study were similar, with a tendency for feedback to focus on adequate performance, highlighting strengths, and neglecting areas that needed improvement, with few action plans included. In addition, written feedback was lacking in many of the forms. Mini-CEX forms have descriptors that guide structured feedback on observed performance,⁸ but these were not used effectively. Possible reasons for supervisors not providing effective feedback could be their multiple roles and responsibilities³³ and the lack of protected time for educational activities,⁴⁰ especially for direct observations or their lack of skills or training.^5,17,39 The supervisors in this study often appeared reluctant to provide honest and negative feedback and grade trainee performance when it is poor, as has been seen in previous studies.^1,5 Other possible reasons for the poor supervisory feedback reported in this study could be deficiencies in trainers’ knowledge and clinical competence, facilitation or interpersonal skills, and awareness of trainee’s learning needs – all factors that have been described in other studies.^2,33 Organisational factors such as work demands, poor recognition for teaching activities and a lack of a conducive environment could also negatively affect feedback provision.¹ Some FP supervisors participated in training courses conducted by the academy or the university but ongoing training opportunities could capacitate them better than once off attending training.⁴¹

In the mini-CEX forms, instead of using competency descriptors, supervisors mainly provided global feedback or non-specific judgements such as ‘satisfactory’, which are vague and do not offer any educational value, as reported previously.^22,23,42 The inadequate space for written feedback on each competency resulted in supervisors recording feedback anywhere on the forms. The WPBA forms can be modified to so-called ‘supervised learning events’, as done in the United Kingdom,⁴³ in which text boxes can replace rating scales with tick boxes to provide narrative feedback. Modified mini-CEX forms with separate spaces to record areas performed well, those that need improvement, and action plans improve educational impact because they act as reminders for supervisors.^36,44 Although general descriptors offer a rough guide for providing feedback, modifying the forms to include the categories ‘competent’, ‘not competent’, and ‘good’ can guide supervisors to provide more specific feedback.⁴⁵

Most FP supervisors scored highly on all mini-CEX competencies and total and global scores across all years of training. The lenient scoring was evident from the first year of training and did not exhibit the regular progression as seen in previous studies.^45,46 The global and total scores progression was evident only between the first and third years, in contrast to previous studies demonstrating score progression across all 3 years.⁴⁶ Ideally, feedback should be provided in the areas where registrars scored low. Insufficient supervisor feedback on the areas that scored low could lead to the registrar’s self-assessing by over- or under-estimating their performance against expected standards.⁸ Feedback should improve and be more holistic during the later years compared with the first year, where it could be limited to fewer competencies and be more specific. This avoids overloading junior registrars with feedback, allowing them to apply it in similar situations and enhancing transferability.⁶ More than half of the mini-CEX forms were not signed by FPs or registrars, suggesting a lack of accountability from supervisors and registrars.

At the end of each clinical rotation, FPs and not non-FPs are expected to assess and provide feedback on whether registrars met the initial learning needs identified in their learning plans. Feedback provided by specialists and medical officers who functioned as educational supervisors may not have been aligned with training expectations. This could have been because of non-FP supervisors being unclear about the registrar’s initially identified expected learning outcomes. A previous SA study also reported that registrars received more feedback from non-FP supervisors, which they found useful.⁴⁷ The role of non-FP supervisors in registrar supervision needs further research.

The reliability of WPBAs increases with multiple assessments conducted at various times by different assessors.^22,45 In this study, registrars had multiple assessors in some districts, but, in others, the mini-CEX forms were completed only by their immediate supervisor. Clinicians functioning as supervisors in clinical settings are neither always trained in medical education⁸ nor equipped to give meaningful feedback.^5,6 This could negatively impact registrar training, even more so in districts where only one supervisor provides feedback. Regular faculty development is needed to understand the educational value of feedback.^22,27 Identifying a core group of well-trained supervisors to provide feedback when conducting WPBAs⁵ could be another strategy to address the reliability of mini-CEX assessments.²⁷

Feedback in the learning plans focused on ethics, professionalism, clinical knowledge and skills and learning behaviours. This was in contrast to a study where professionalism was the learning goal least identified by trainees.⁴⁸ Feedback provided by specialists and medical officers was much more comprehensive, covered many competencies and was aligned to registrars’ learning needs (Figure 2). Supervisors’ ability to provide feedback on appropriate history-taking and physical examination skills is critical in workplace learning, as clinical diagnosis depends on these registrar skills.⁵ Mini-CEX feedback mostly covered history-taking, physical examination, comprehensive management and organisational efficiency. Clinical reasoning was the least covered. This differed from previous studies where feedback focused least on professionalism and organisational efficiency.⁴⁹ As reported previously, there was no alignment of many feedback areas to the registrars’ learning needs.⁴⁹ Family physicians provided scanty feedback on soft skills, such as clinical reasoning or patient negotiation skills (Figure 3). High scores given by FP supervisors on all competencies suggested insufficient skills to evaluate registrars critically, especially for soft skills. Ideally, more narrative feedback should be offered on identified learning needs where supervisors gave low scores.

This is the first sub-Saharan study to explore the significance of training sites and supervisor characteristics on the quality of feedback. While one district (D3) had fewer FPs involved in training, registrars received more comprehensive feedback than other districts with more FPs. This infers that feedback does not depend on the number of supervisors but is context-specific, dependent on willingness and effort, understanding of feedback characteristics⁵ and assessment literacy.^7,22 All supervisors and registrars involved in decentralised clinical training require ongoing faculty development training on what constitutes adequate feedback to maximise this essential element of workplace-based training and assessments in decentralised settings. There are several training options that could supplement the existing training. One option is to conduct practical sessions using videos of actual registrar consultations and the supervisory feedback that could be critiqued by pairs or groups of FPs as done in other settings.⁵⁰ Another strategy could capitalise on the peer coaching or co-teaching¹ opportunities provided by observed consultations – pairs of junior and senior FPs should provide feedback to registrars after these sessions. Similarly, feedback sessions between FPs and registrars in combined sessions with medical students or interns at clinics or hospitals provide opportunities for registrars to practise and internalise effective feedback. Finally, and most importantly, FPs should promote an empathic learning culture based on constructive feedback, even at the lowest family medicine training level, at the primary healthcare facilities.

This study evaluated supervisors’ feedback in family medicine registrars’ LPs. Supervisor feedback was shown to be inadequate, often very general and not very helpful in developing action plans to improve registrars’ skills. The lack of detail tailored to individual needs could negatively impact the training of health workers. Family physicians’ supervisory role is vital to ensure the training of skilled family physicians to strengthen primary healthcare and district health systems, which should translate to better health outcomes in communities. Future research on feedback in postgraduate family medicine training programmes in similar contexts could help to improve the quality of this cadre of health professionals.