When systems set the limits of supervision

Abstract

Supervision in postgraduate medical education is often described through its component parts: feedback, observation and entrustment.¹ These are the practices we return to when we think about improving training. They are visible. They can be named, taught and assessed. Across the literature, they are also consistent. And yet, the learning that emerges from them is not.

Wiese and Bennett offer a way to make sense of this.² Their study shows that these practices are present across clinical settings, but they do not come together in the same way. What differs is how they connect within the conditions of everyday work. Continuity, workload, team structure and local culture shape how supervision unfolds. Over time, these elements settle into patterns. Those patterns, more than any single practice, shape what learning becomes.

Seen through a systems lens, that variability stops looking like a problem to be fixed and starts looking like evidence. Different conditions produce different configurations. Different configurations produce different learning. The variation is the system showing you how it works—and what it makes possible and what it does not. Those differences across training environments are not only variations in how supervision is enacted. They also reflect the limits that systems place on what supervision can achieve.

That ‘systems’ framing changes how familiar challenges appear.

When trainees struggle to progress, attention often turns to strengthening individual practices. Feedback can be made more specific. Expectations can be clarified. Observation can be increased. These are important responses. They reflect a genuine commitment to improving training.

At the same time, research has shown that practices such as feedback are shaped by relationships, credibility and the broader environment in which they occur.^{3, 4} Decisions about entrustment unfold within uncertainty and are responsive to contextual cues as much as to trainee performance.^{5, 6} Supervisory practices do not operate on their own. Their effects depend on how they connect to other elements of the system.

Feedback carries weight when learners have an opportunity to act on it.⁷ Observation becomes more meaningful when it occurs within relationships that provide enough continuity to make sense of what is observed.⁸ Support seeking contributes to learning when it opens space for dialogue and shared problem solving.

When these connections are in place, learning can move forward. When they are not, the same practices remain present, but their impact is diminished.

Context has long been recognised as important in workplace learning. A systems perspective takes a further step. It helps explain how context shapes what supervision can actually accomplish. It brings into view not only variation, but the boundaries within which supervision can function as a developmental process. Some systems support progression in sustained ways. Others make progression more difficult to maintain. In some, connections between practices are harder to hold, and learning becomes more fragmented. What is striking is that these are not gradual differences in quality. Systems can appear to function adequately, even robustly, until the conditions that allow practices to connect are sufficiently disrupted. At some point, the issue is no longer the quality of individual practices. The practices are still there. What has changed is their ability to connect in ways that support development.⁹ These are the limits of what the system makes possible and understanding where those limits lie matters as much as improving what happens within them.

And yet, knowing where the limits are does not mean people cannot still work within them. Systems are enacted through everyday decisions—and those decisions are shaped by what educators notice and what they do not.

Educators decide when to step in and when to step back. They interpret what they observe and respond to uncertainty. Those patterns are not always visible to the people acting within them—and experienced educators often seem to sense when something is not quite aligning. When feedback is present but does not lead to change. When close oversight is not translating into greater independence. When support is available, but learning is not deepening.

These moments are often described as examples of good teaching. They can also be understood as moments when the system itself becomes visible.

If learning emerges from how practices connect in context, then expertise involves more than performing those practices well. It involves recognising how they are interacting and responding accordingly. That response is not about adding new practices. It is a different orientation towards the ones already in use: reading how practices are connecting in the moment and responding to what that reading reveals.

This capacity is not commonly named in medical education, yet it is implicit in accounts of experienced clinical educators. It echoes broader understandings of professional expertise as involving pattern recognition and situational awareness.¹⁰ Here, that orientation turns towards the supervisory system itself: not only reading clinical situations but reading how learning is taking shape within them.

This capacity does not remove structural constraints. Where continuity is limited, where workload restricts interaction, where relationships are difficult to sustain, the conditions for learning are already constrained. There are limits to what individual educators can change. But what is recognised in those moments still matters. It shapes how educators respond, what they prioritise and how they make sense of what is happening around them.

Efforts to enhance training have often focused on ensuring that key practices are in place and that educators are supported in developing them. This has led to important advances in feedback, assessment and programmatic supervision. A configurational perspective adds a different layer to this work.

Improvement begins with attention to the conditions that allow supervisory practices to connect in ways that support learning.^{11, 12} It also depends on recognising when those connections are weakening, absent or producing unintended consequences.

Wiese and Bennett have provided a compelling account of supervision as a contextually configured system.² The next step is to reckon with what this means for how we approach improvement. Until we attend to both how systems are designed and how they are perceived in action, we risk continuing to ask educators to optimise practices within systems that quietly set the limits of what those practices can achieve.

No conflict of interest.

Not applicable.

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.