Diagnosing the learner: Tips for teaching struggling medical trainees

Medical training is riddled with transition points, from medical student to intern, intern to senior resident, then resident to fellow or attending. Most of us who choose careers in medicine are prepared to be lifelong learners. After years of higher education, learning is a natural part of our lives. We know how to do that! The transition from learner to teacher is less talked about, and one I felt much less equipped for in residency.

I am currently a PGY-3 internal medicine resident at the University of Washington in Boise. In my role, I frequently oversee third-year medical students completing their internal medicine clerkships, fourth-year Sub-Is and interns on the inpatient medicine wards. As a senior resident, I found my new role of “teacher” to be manageable, even fun, when my trainees did not need excessive help. But when they were floundering, I felt a little lost in my efforts to get them caught up. What do you do when you are working with someone who cannot seem to put the big picture together? How do you help someone who is struggling to get to that next level in their clinical practice?

It turns out, as I started speaking with colleagues and looking into the literature, I was not alone in asking these questions.  In this article, we will go over some tools you can use to diagnose your learners so you can give more specific, constructive feedback at all stages of training.

Evaluating mental health and competency areas

When evaluating a learner who is struggling, there is one “can’t-miss” diagnosis that should always be considered: mental health. An estimated 40% of struggling learners are dealing with an underlying mental health issue or learning disability, most frequently anxiety, depression or undiagnosed/inadequately treated ADHD. This is the elephant in the room that must be addressed before turning our attention to other domains.

Unfortunately, there is no perfect formula to address such a personal issue in the workplace. I would encourage teachers to be thoughtful about when and where this is addressed. While rounds might be where you usually interact with trainees, it is certainly not the place to voice these concerns. Carve out a specific time for bi-directional feedback—be sure to choose a private, neutral setting for this.

Reach for open-minded questions like, “How’s life outside of medicine?” or “Medical training is tough; how is your mental health holding up?” Remember that these are questions we should ask all our learners, not just the ones who are obviously struggling.

If we feel reassured that mental health is not the central issue, we can begin to thoughtfully evaluate trainees by focusing on four major competency areas:

1. Medical knowledge
2. Clinical reasoning
3. Organization and efficiency
4. Professionalism and communication

While not all-encompassing, most learners who are underperforming clinically are experiencing a breakdown in one of these four domains. The most common primary deficits are in clinical reasoning or organization/efficiency; the former will be the primary focus of this article. We will specifically look at how to evaluate and troubleshoot deficits in clinical reasoning. For learners who are struggling, it is important to figure out where in the clinical reasoning pathway things are breaking down.

The bare bones of this pathway include: 

1. Formulating an initial impression of the patient’s problem
2. Data collection
3. Problem representation
4. Refining the hypothesis
5. Selecting a working diagnosis

This process is not always linear in real time, with experienced clinicians often mentally working through several of these steps simultaneously while evaluating patients. We will review each step in more detail below and identify where problems may manifest.

Breaking down the pathway

Forming an initial impression/early hypothesis helps focus the history and physical aspects of a patient visit. Breakdown in this step may manifest as inefficiency or disorganization if the learner is not sure how to filter key data and is getting bogged down in the details. Think about the learner who arrives early and is one of the last to leave, but still does not seem to have a good grasp of what is going on with their patient when presenting.

Despite a good medical knowledge base, they may regurgitate an excess of data when presenting on rounds or conversely, leave out parts of the history that should have helped guide the diagnosis. When this occurs, go back to bedside together. Model any important or diagnostic exam maneuvers. If their history contained gaps, fill those in at this time. You cannot confidently develop a working diagnosis if information is missing.

There are a few exercises that can help trainees practice hypothesis-driven data gathering. One is called “searching for the scripts,” which is similar to the game “20 Questions.” Provide your team with a chief complaint and have them give a list of three possible diagnoses, prioritized. They can then only use five pieces of the patients’ history and five of the physical exam to differentiate the possibilities.

You could similarly assess this by printing a history and physical for your learner and having them highlight only the details that help to narrow the differential.

Tackling problem representations

You may have interacted with learners who seem to “miss the big picture” or were not able to “put it all together.” This feedback, while frustratingly vague, alludes to a breakdown further in the clinical reasoning pathway: difficulty with problem representation. It may be especially apparent during handoffs, consult calls or other oral presentations when a trainee needs to be able to communicate the highlights of a case in a brief, meaningful way. 

The summary statement, or “one-liner,” should be accurate, up to date and succinctly piece together the key features of the patient’s case. Using semantic qualifiers such as acute vs. chronic, intermittent vs. constant, proximal vs. distal, etc., is associated with strong clinical reasoningand should be included in the problem representation. This drives the differential diagnosis by triggering appropriate illness scripts.

If a learner is struggling with the one-liner, deconstruct the problem representation into its three main components:

1. Patient demographics and risk factors: think about gender, age, relevant exposures/substance use, predisposing conditions, etc.
2. Temporal pattern: this is where those semantic qualifiers come in
3. Key symptoms and findings

As a teacher, be sure to articulate your own problem representation to model what you expect from your learners. Share your reasoning aloud, especially when linking pieces of the summary statement to your illness script. 

The hypothesis generation and differential building

Good clinicians should be able to generate a wide differential, even with limited information. A learner who produces numerous possible diagnoses for a given case but is unable to prioritize their list may not have formulated solid illness scripts for the diagnoses being considered. Have the trainee compare and contrast the diagnoses and explain their justification. Review the illness scripts for their top diagnoses and see what fits best. Do we need any more pieces of data (ex: different labs, imaging, etc.) to discriminate between the top contenders?

Students often receive vague feedback to “read more,” but we can refine this to be more specific and constructive. Encourage useful reading habits by having learners read about patients in a way that promotes diagnostic reasoning. One way to do this is to have the learner always read about at least two diagnostic hypotheses at the same time (example: gout vs. septic arthritis). This helps the learner pay specific attention to the discriminating clinical features of a disease, which become “anchor points” in memory.

Diagnosing both the patient’s clinical problem and the learner’s ability is a challenging art form. In the future, when your learner is struggling, ask yourself:

• Could there be a mental health issue at play?
• Where in the clinical reasoning pathway are things breaking down? (Data gathering, problem representation or differential building/triaging)
• Does the learner have a good sense of the illness script for this condition?
• Are your expectations reasonable for their level of training/clinical experience?

We are all learners and teachers at some point in this journey. I hope you have some more tools at your disposal now to help your trainees hone their clinical reasoning skills. Good luck and happy teaching!

Related reading:

Weighing your options: Should you pursue a research year?

A guide to medical terminology