From the outset, there was a clear understanding of the importance of the context of learning, which meant that the focus shifted from the university campus to the practice locality of the doctors. While students were required to attend a 5-day course annually at the university, the focus was on the context of the students who were located all over South Africa, mainly in rural hospitals, districts and practices. However, much of the learning happened in the practice context, supported by regional seminars to help students reflect on and change their practice, and to build local networks. In recent times, decentralised or distributed medical education has become very common,^2,3,4 and the importance of context has also been emphasised.^5,6,7 Yet, Fehrsen was standing on the shoulders of others in grounding the MPraxMed in the context of his trainees. Freire had challenged the way that educational institutions colonise thinking so that students come to believe that their lived experiences in context are irrelevant to learning, whereas these are the basis of critical consciousness and action.⁸ Fehrsen would certainly have agreed with Gruenewald, who proposed a critical pedagogy of place, which challenges educational practices that disregard place,⁹ to focus on more socially and ecologically just approaches. Reid, a graduate of the Medunsa MPraxMed, has argued that this provides a theoretical framework for a distinct rural pedagogy.¹⁰

For many years, there was a stream for full-time family medicine registrars doing the MPraxMed who were located within the teaching hospital complex, but this was phased out because the context of learning was too far removed from primary care in the periphery. Although this was commonly how family medicine training was provided internationally, Fehrsen came to believe that it was difficult for students to learn the principles and practice of family medicine when working in specialised departments or sub-departments in large, central hospitals, where the whole ethos is biomedical, fragmented and disease-centred rather than biopsychosocial, continuous, patient-centred and community-based. This does not mean that specialists did not, and do not, have a role to play in teaching appropriate skills, but they should not control what, where and how Family Medicine is taught.

Here is an example of how this shift influenced the course. In 1997, doctors from three rural hospitals in one province, all generalists, enrolled en masse for the course. Obviously, they could not all leave their hospitals at once. Thus, the first 5-day course was held in one of these hospitals, about a 6 hour drive from Medunsa; at the first session, there were 18 MFamMed students as well as non-student doctors from neighbouring areas. An anaesthetist and an obstetrician accompanied the Medunsa team. A lot of learning took place in the wards and theatre – the hospital kept on working while the course was happening, with everyday problems providing learning material. The then medical superintendent of this remote rural hospital wrote:

This method went on to be repeated in other regions.

The foregoing example helps to make the point that the value of this shift lies not only in the learning that takes place, but perhaps even more in the value given to the context, validating students’ practice sites as places for everyone’s learning and growing.

Learning from one another and not merely from teachers was seen to be essential. Teachers like to teach. However, what is taught is not necessarily learnt, and certainly not often applied, as we know very well from patient care. Thus, the focus was shifted to learning: how and why do we learn, and how can educators help learning to occur. An important part of the MPraxMed course involved helping students to reflect on their learning process. Education literature now commonly refers to self-regulated learning (SRL)¹¹ and self-directed learning (SDL),¹² both of which include the learner’s active engagement in goal setting and choice of learning strategies as well as the key role of reflection, and which are often used interchangeably.^13,14,15 At the time we^* used the term self-initiated learning, which can be seen to fall under SDL,¹⁶ because we focused on the motivation that was expected from our students – the course depended on their learning initiative, because we came to realise that if doctors rely upon teachers, they would be stranded as soon as they finished the course, whereas the aim was to provide skills for ongoing learning and adaptability in ever-changing situations.

Furthermore, we recognised that students learn much from each other, and are often more open to learning from peers than from us, because they feel their peers know exactly what they feel and face every day. Peer (and near-peer) learning in clinical education is increasingly common and has been shown to be beneficial.¹⁷ Clinicians learn from each other all the time, so there were few challenges experienced in using this approach. During training sessions, students in all 3 years of study were mixed during whole group sessions as well as small group discussions, to assist with this learning process. It was an important affirmation for MPraxMed students to discover how much they actually do know, especially in relation to the rural contexts they work in, and also how much they can learn from each other.

Historically, learning opportunities were determined by what ‘the experts’ thought students needed to know. In the MPraxMed programme, we sought to let the students identify their own learning needs. As part of that, we became uncomfortable with the term ‘teacher’ or ‘lecturer’, and moved to becoming facilitators of learning. The concept of facilitation of learning in health professions education has mainly been described in the context of nursing education, where it has been argued that students need to develop confidence in their knowledge and skills to translate these into effective practices, which requires them to explore and question rather than passively accepting and modelling, an approach that emphasises the experience of the learner and their individual needs.¹⁸ This describes well the approach that we adopted, as we strove to hand over control to our students, to let them express their learning needs and agendas, and to work with them in seeking to meet those needs. As part of this, we too sought to move towards being better facilitators, giving feedback to each other and seeking it from our students.

This required individual and group negotiation, but ultimately, it was our intention that every training session should be co-created with students. Students chose their own topics for module assignments, according to the problems they encountered in their practices, and these were assessed on the degree to which the students’ own learning needs were met. For specific year group activities, students in a particular year planned, according to the curriculum guidelines, which themes they wanted to tackle together and in what way.

In small group sessions, the role of the facilitator was to ensure the learning process was as effective as possible, giving input as a fellow group member. This could be frustrating, and the urge to intervene in order to teach was often great, but we learned to trust the group process. One experience was starting off with a new first year group, whose learning process consisted of laboriously reading to the others sections from pre-readings which their colleagues had also read, which left the facilitator (IC) feeling quite desperate! But after reflecting on a couple of sessions such as this, the group itself began to say, ‘Something is wrong; this is not very interesting or educational. How can we do things differently?’ Slowly, the process transformed so that readings were summarised and applied by group members, and fruitful discussions were held on their relevance and applicability. This also helped to develop critical reading skills.

The pull to control was so strong that we failed at times, especially with students who had previously learnt in a system that encouraged spoon-feeding and rote learning. It was particularly difficult when students did not respond to our attempts to share control and found it difficult to participate in this new approach. Regular reflection sessions among the facilitators assisted us to stay on course and for us to become more skilled at (and comfortable with) facilitation.

At the outset, the MPraxMed offered fixed courses. Steadily, we began to offer a range of options and resources to enable students to satisfy their learning needs, to become better doctors for their context.

Students were presented with a number of themes, which formed the content of the programme, such as Principles of Family Medicine, Whole Person Medicine, Ethics, the Family and Community, Culture, Research, among others. The themes had been presented as courses prescribed for particular year groups. Subsequently, students were provided with copies of core articles for each theme and lists of alternative references, advised regarding core and recommended books, and, increasingly, pointed to online resources. Students then decided, as individuals and in their learning groups, what they needed to address and in what way. This was pre-Google, so relied on a well-managed Resource Centre where students could find relevant material.

This ensured that instead of presenting information that was already known to students, or issues that were not relevant to their practices, they focused on what they needed to know at a particular time, and were thus much more open to learning – they could scratch their itches, not ours! This went beyond the flipped classroom concept, which came much later, but which has proven itself to be more effective than lecture-based learning.¹⁹

For new facilitators, this could be a scary process. It is much easier to prepare and ‘give’ a course! IC vividly remembers his first encounter with a newly enrolled group of students, who came expecting to be educated in the traditional way. When he threw the ball back into their court, he was not sure who was more anxious – the students as they tried to deal with this new concept of a lecturer who actually wanted to know what they needed to learn, or IC as he desperately hoped the students would come up with a topic that he knew something about!

To assist students to meet their learning needs, we first have to hear what these needs are. It is an interesting process to monitor the airtime in a group; in a traditional group, the leader or teacher will obviously dominate, whereas we were learning to take more and more of a backseat, to ensure the students had more airtime. It was a vital task to try to hear and understand our students and their needs in order to help them better.

This is, of course, directly analogous to a patient-centred consultation, and that is what we wanted to model. Fehrsen wanted our principles of teaching and learning to be congruous with our principles of consulting. So just as we aimed to be patient-centred, we aimed to be learner-centred, which implied actively listening to our students the way we actively listen to our patients. The importance of role-modelling in medical education is widely described in the literature, and is credited with integrating beliefs, values and principles into a doctor’s worldview, thus affecting professional identity formation.²⁰ It is acknowledged that these effects depend on contextual, structural, cultural and organisational influences, including the relationship between students and preceptors.²⁰ More significantly, it is argued that imitation may perpetuate undesirable practices, such as doctor-centred patient interviewing, which requires student-centred approaches to mitigate against this.²¹ There is scanty literature on the link between student centredness and patient centeredness, with only one study being found,²² but the parallels between learner–teacher and doctor–patient relationships have been observed,^23,24 and we believe that there must be some carryover.

Obviously, we failed in this approach many times. We often felt we knew so much and desperately wanted to share all the information at our disposal, but as soon as we did that, we found that we stifled our students’ growth, just as we have seen happen when we dominate in our patient interactions.

We recognised that we too were on a path of learning. Although we had additional experience and knowledge acquired over the years, these may not have been relevant or helpful to students. We were co-learners in a number of senses:

In essence, once again we were modelling a way of being to our students that has implications for our relationship with our patients. Fehrsen’s philosophy was that as soon as we pose as experts, we not only set ourselves up for failure, but we might also reach a dead end where we stop learning, sit back and say we know enough.

In our early days as facilitators, we worried that students might ask us questions we could not answer, until we realised that it was putting ourselves back into the role of expert rather than co-learner.

The starting points of learning were the patients and the practice of each student. One of the key roles of the facilitator was to keep the students focused on their own patients and their own context. We believed that patients are the most valuable teachers, and that every patient interaction provides an opportunity for learning. Thus, patient interactions were what we sought to get our students to focus on.

Working within the structure of a university programme, assessment was required. Despite our best intentions to make the assessments relevant to practice, there was a tension that students could become more concerned about the examination than about improving their practice. We recognised that assessment can have unintended negative consequences for learning,²⁵ although we did not at the time have the current language of assessment of learning versus assessment for learning.²⁶.Intuitively we knew that we needed to consider the broader environment and culture in which assessment was taking place,²⁵ and to try to ensure our assessments were driving the right kind of learning.²⁷ We had introduced a form of more workplace-based formative and summative assessments, in that students were required to bring video recordings of real consultations from their clinical practices, which were very useful, despite technical challenges. We also started basing regional contact sessions near clinical facilities at which students worked, so that students could spend a few hours observing each other in consultations and offering feedback to their peers (an early form of workplace-based assessment), and we started practice visits to students’ own places of work, both as a mentorship and assessment tool. However, our assessment remained an element of incongruity in the course, which we struggled to resolve.

Towards the end of the 1990s, we were trying to move students from acquiring theoretical knowledge to improving performance and practice through reflection. This was based on an understanding that the actions of an expert arise out of a complex integration of knowledge and skills that are appropriate to the particular context they are in and the situation that they face,²⁸ and that reflection is the process of analysing and learning from one’s experiences to improve professional performance. This requires both reflection on action, which is reviewing a decision or event and considering all the factors that impacted the outcome, and reflection in action, being able to consider these same factors in real-time to modify a decision in order to improve an outcome.²⁹ We began to build application of new knowledge and reflection on this action into all tasks and assignments. We were realising that knowledge is constructed through action and participation and were seeking to develop this theme further. We felt that if a student obtained the degree, but their practice was unchanged, then we had failed – the overarching aim was to strengthen care for patients.

During his own training, HC experienced a profound shift from being doctor-centred to becoming truly patient-centred. Patient-focused learning became real through his patient studies, each based on individuals he encountered in his everyday work. He still remembers many of these patients, not just their diagnoses, but their stories, their faces, and the lessons they taught him. This transformation also influenced his approach to teaching, making him more student-centred in his role as a facilitator of learning. His own lifelong learning, as well as the way he guides students in their learning journeys, continues to be deeply rooted in his patients.

Although there was much less health professions education (HPE) literature available at the time, looking back now with an understanding of the evidence, we believe that the MPraxMed was already demonstrating a number of the approaches to learning that the evidence now urges educators to use,³⁰ foreshadowing the role family medicine plays in driving change in medical curricula, both undergraduate and postgraduate, in South Africa and globally.³¹ An example is the development of a capability approach to learning.³²

Postgraduate family medicine training in South Africa has continued to be at the forefront in implementing a number of these principles, including the critical importance of context, distributed learning and the value of patient centredness in training. This is not surprising given that at the time of Fehrsen’s death, five heads of the eight Departments of Family Medicine in the country were graduates from this programme.¹

Facilitating learning by students who are empowered to regulate their own learning and for educators to be co-learners with our students requires reflection on and monitoring of our identity as educators. Flattening the traditional teacher-student hierarchy is still found to be threatening by many colleagues, especially because this approach does not always find acceptance among their teacher peers. At the same time, artificial intelligence is reinforcing the importance of moving from a focus on gaining knowledge to developing the ability to evaluate and use resources wisely. Assessment is moving strongly into the workplace. It is heartwarming to note the extent to which family physicians are at the forefront of driving workplace- and competency-based assessment in registrar training in South Africa. We urge our family physician colleagues to courageously continue to use and develop best educational practices with registrars in the hope that they too will develop not only patient-centred passion but also excellence in how they teach future generations.

The authors were part-time facilitators on the programme in the 1990s; three were also students on the programme preceding their role as facilitators (HC, JB and IC). We embrace our subjectivity as well as the unique insider insight that comes with that background. All authors have subsequently taught on other Family Medicine Masters programmes, have been engaged in education more broadly and have played leadership roles in Family Medicine in South Africa, all of which afford us unique perspectives from which to reflect on our experience.