One of my colleagues recently referred to their RMO (resident medical officer) as their random medical officer. This was no criticism of an individual but reflected the developments over the last decades where knowing exactly who’s in one’s clinical team has become less certain. There was a time when, in hospital settings, a medical team comprised a senior doctor, registrar, RMO and sometimes a trainee intern—more importantly, that team remained relatively stable for weeks or months. They all got to know each other, they learned each other’s strengths, weaknesses and foibles, and apprenticeship learning just sort of happened. Let’s not forget the price of that continuity—long, unreasonable and unsafe working hours. Those days are gone. In this issue of the journal, Rassie wonders if the days of the old apprenticeship model are also gone.1 She calls for a structural change to the apprenticeship model. There certainly needs to be a rethink of apprenticeship—I’m less sure a structural change is the solution. In either case, both she and my colleagues seem to agree that apprenticeship isn’t quite right at the moment.

The pendulum is swinging around apprenticeship. It was once seen as the cornerstone of good medical education. It then fell into disfavour and was viewed by some as inefficient, at times exploitative, and often as a way of preserving the status quo—a way of ensuring “the way we do things around here” stays that way. Evidence-based practice, standardisation of education and wanting to be sure all learning objectives were learnt led to a period where apprenticeship was seen as too messy and uncontrolled. Then we rediscovered workplace learning and with that, began to understand what makes it work.2–4 Here’s my view of what makes workplace learning work.

Let’s start with supervision. It’s easy to get this wrong. We can over-supervise or under-supervise—both have problems. When we over-supervise, we take too much control. We don’t allow the trainees sufficient autonomy to make their own decisions. We end up just telling them what to do. This creates passivity in the trainee and they can become too scared to make any decisions without deferring to their supervisor. This just makes the supervisor even more controlling, and a vicious cycle is soon in play. When we under-supervise, we leave the trainee to it—“call me if you’re worried” with the subtext of “don’t really call me or if you do I’ll consider you incompetent”. The trainee may well learn a lot—often by making mistakes—and there are real concerns around patient safety. Neither of these scenarios is great for learning.

But judging the right amount of supervision to provide is also difficult because every trainee is different—made worse if there’s a new trainee on the team every week or two. Likewise, every supervisor is different—each has his or her idiosyncrasies. As an example, I recall the wall of an orthopaedic ward covered with notices, each explaining how each orthopaedic surgeon liked their DVT prophylaxis to be given. None of the protocols was wrong, but they were all different. If a trainee did not know that and applied a different protocol, they might be chastised for doing it the wrong way. If they asked the surgeon how to give DVT prophylaxis they may fear being seen as ignorant—“didn’t they teach you DVT prophylaxis in medical school?”

Herein lies the heart of the problem. Each trainee is different and each supervisor is different. Yet, if these are not understood by each party, how can trust develop? How can a supervisor know what it’s safe to let a trainee do? How can a trainee know what they’re allowed to do and what they should ask about? In the old days, people just worked it out because they had time to. Yes, the first weeks were tricky while each party got to know each other, but eventually it all settled down and a working and learning relationship ensued.

I suggest therefore that part of the solution to the new apprenticeship is a form of speed-dating. We need ways by which learner and supervisor can quickly get to know each other and quickly learn each other’s strengths, weakness and idiosyncrasies. This establishes mutual trust and is a process change, not a structural change.3 What might this look like? It may be taking a few minutes whenever a new person joins the team, and call me old-fashioned here, to get to know each other; attending to the initiation part of the team building, not just doing the maintenance part.2 This needs to be two-way—making space to learn about the trainee’s strengths and weaknesses as well as making space for the trainee to learn about the supervisor’s strengths and weaknesses. This includes making it easy for each other to ask questions4—“there’s no such thing as a stupid question in our team”. Rassie notes that assessment may undermine this1—that may well be true, particularly if assessment aims to judge someone’s knowledge, but what if assessment were to judge someone’s curiosity? Suddenly, assessment might actually help—another simple process change. This speed-dating needs to occur every time a new person joins the team—for some, this may be every week. That would take time but we’d probably get better with practice.

It’s easy to get trapped into equating good supervision and good apprenticeship with good teaching, and then to think that good teaching equates with telling people what to do. Workplaces rarely work that way—they’re often too busy. Yet, this busy-ness is its strength. We recently undertook an ethnographic study of learning in a ward where two of the interesting findings were that the learning moments were often very short—the median duration was one minute with a range from 15 seconds to 21 minutes. Secondly, many of these moments went unnoticed.5 So, learning is occurring all the time, we just need to make it explicit.

Good apprenticeships foster good learning.6 Of course, we also want good teaching and good assessment. But for me, I’d start with a focus on ways to form quick and effective supervisor-trainee relationships within the first few minutes of someone joining the team—from that everything else follows.