Providing feedback and defining excellence in medicine

There appears to be wide agreement that we are not good at providing feedback in medicine. I have attended many seminars and read many papers on this topic, and despite believing this is an essential role in medical education, I still fail routinely.

I have received a lot of advice. I should start by ‘setting the stage’, picking an appropriate time for feedback, and explicitly flagging that I am about to provide feedback. I need to collect evidence, and provide feedback based on objective observations. I need to first establish trust. I need to explore the learner’s perspectives before expressing my own. That is all excellent advice, but providing good feedback remains incredibly difficult. 

After 15 years struggling with this issue, I realized there is an essential first step that was left out of almost every discussion. Before you can provide effective feedback, you must first define good. How can you move your learner towards being a better doctor if you haven’t first defined the ideal doctor in your head?

When talking about feedback, people often say they are nervous about giving learners negative feedback. However, I know doctors, and most doctors love to speak their mind. I don’t think most of us are truly nervous about having difficult conversations. Doctors have difficult conversations every day. When there is an obvious gap (you are holding the laryngoscope in the wrong hand), we don’t hesitate for a second before correcting it. (Although one major problem is that good feedback requires us to actually observe our learners. Modern emergency medicine is rarely set up for us to have time to watch learners performing routine tasks, like histories and physicals, and the most important part of medicine – the clinical reasoning – cannot be directly observed.) Personally, I think we are comfortable providing negative feedback when it is required, but we often don’t know what to say.

By the end of a shift, I generally have a sense of whether a resident is ‘strong’ or ‘weak’, but I often have a hard time articulating why. I have a gestalt comparison to prior learners. Much like the end of the bed diagnosis of DKA, I can sense that this learner is likely to struggle or excel, but I have a hard time articulating why. How am I going to provide excellent feedback if I can’t articulate my thoughts? How can you guide learners towards being a better doctor if you don’t have a definition of the ideal doctor framed clearly in your head?

I am not an expert in feedback. I don’t think you should take my advice. Many, many people have written about feedback. There are many different models you can use. Learn the specifics from someone else, but I wonder if others have run into the same wall that I have? Before we can even start providing feedback, we must first clearly define the desired result; the ideal physician. 

In my mind, a good feedback conversation probably consists of 3 discrete steps. The first has nothing to do with feedback at all, at least as we normally envision it. The first step is a discussion about the definition of good. In medicine the question is “what makes a good doctor”, but I think this can be applied in many other contexts. If you want to have a good conversation about your relationship, before jumping into ‘feedback’ or talking, you better have a clear conception in your mind of what makes a good partner or a good relationship.

Because there is no single agreed upon definition of the perfect physician (or perfect relationship), you would ideally engage in a discussion about this goal before ever considering feedback. I am not an absolute relativist. I believe that there are good and bad ways to practice medicine. I am not saying that everyone gets to decide on their own definition of “good medicine”, but a feedback conversation is unlikely to be productive if you don’t have some shared understanding of that goal. 

The second step should be pure observation, without judgement. “Here are the things that I observed that either matched the criteria of good or didn’t”. You want to approach this with curiosity, or a true desire to understand, and let the learner tell their side of the story. 

Example: “We agreed that being a great doctor involved using a thorough history and physical exam to generate a complete and accurate differential diagnosis for each patient. When you presented Mr. X, you didn’t describe the rash that I noticed on his back, and left shingles off your differential diagnosis. What happened there?” The answer might be a knowledge gap (ie, I didn’t think to look), or it could be something else altogether, such as “I was seeing the patient in a hallway bed and was uncomfortable exposing them in front of the entire department”. Curiosity is important, because the appropriate feedback differs dramatically. 

The third step is to find a path to close the described gap. Sometimes this will be prescriptive (“I think you should do x”), but it is probably more effective when it is collaborative (“How do you think you can improve for next time?”). The exact advice will depend on the situation, and will vary depending on the answers given in step 2. Honestly, this part is pretty easy.

The hard part of medicine is generating the correct differential diagnosis. Once the differential is sitting there on paper, the best course of action is usually pretty obvious. The same is true in feedback. The hard part is identifying the true objective, and the gap between current practice and that objective. Once the objective and gap are identified, the course of action is usually pretty obvious.

What is the ideal physician?

Now we come to the hard part of this article. We need to define a good doctor. In order to have a reference stand that I can push my students towards, I need to have a clear definition of medical excellence. Unfortunately, sitting here, I realize that I don’t. At least, nothing explicit. I have some vague concepts, as I imagine many of us do, but I don’t have a clear definition, which is probably why I suck at feedback. 

I am really hoping this post will be interactive. I am going to throw out a few characteristics that might be important in an excellent physician, but I hope people will use the comments section to help me to refine my definition over time. In the end, my hope is that the discussion helps all of us provide better feedback by allowing us to clearly set our compass, and therefore better identify when and how a learner might be off course.

An excellent physician is:

• Compassionate. Someone who cares about the patient as a person, not just a medical mystery.
• Humble. They will understand their limitations, and know when to ask for help, or turn to resources for support
• Thorough. They will take the time it requires to hear the patient’s entire story, and consider all possible explanations (create a full differential diagnosis).
• Curious. Curiosity leads to more questions, and better understanding of patients. It leads to lifelong learning. It leads to improvements not just of the individual, but perhaps medicine as whole. 
• Driven. A good doctor wants to improve. A good doctor will seek feedback. They will collect not just opinions from colleagues and mentors, but statistics on themselves that they can compare against their peers. 
• Scholarly (or perhaps “scientific” is a better word). Most medicine is practiced outside of academia, but I think great medicine requires science and learning. This is related to “humble”, “curious”, and “driven”, but I think it highlights the importance of the scientific method to medicine, and allows us to point out the flaws in quacks who prescribe ivermectin or advise against vaccines who might meet other criteria of excellence. 
• A good communicator. Do I need to expand here?
• An advocate. Medicine is not perfect. Improvements are necessary. Segments of our society get left behind. A great physician cares not just for the patient in front of them, but for the entire community. 
• Professional. This is not just a job. It is something more than that, and our behviour needs to reflect that fact.
• Prudent. Doctors need to be cautious and detail oriented. This may overlap significantly with being thorough, but it also comments on the way we make decisions. Patients don’t want us to take risks (although they also don’t want us to be overly cautious). Being cautious in our work and in our decision making is important in medicine.  
• Expert. I put this lower on this list on purpose. Long thought to be the bedrock of medicine, I do think expertise is essential. We need to have both breadth and depth of knowledge to help our patients. However, I struggle with the exact definition and role in feedback. Learners are there to learn, and therefore aren’t expected to have all medical knowledge memorized. No physician can be expert in all fields, which is why humility, curiosity, and learning are all higher on this list. However, there is some baseline knowledge that is absolutely essential to perform this job well.

Some criteria will be more specialty specific, but still important to understand. For example, a good critical care doctor will care deeply about mastering procedural skills, breaking procedures into their smallest components, and working to eliminate or minimize all potential sources of harm. So perhaps I need to repeat this process, with a more specific definition of the excellent emergency physician. 

Further thoughts

I don’t really think there is a single definition of good. Most things in life occur along a spectrum, and there are often virtues to be found in opposing views. However, as I said, I am not an absolute relativist, and I believe that there are some core skills all physicians need to excel. 

In medicine, I think the important thing is to be explicit about your values. If I am teaching in emergency medicine, my specific values might be different than those of a dermatologist, and I might need to keep that in mind when teaching off-service residents. Indeed, I know that within emergency medicine, many people will argue with me about the definition of excellence. That is fine. I think we just need to be a lot more explicit about our definitions if we are going to personally strive towards excellence, and if we are going to try to push or pull our learners in that direction as well.

Assessing and providing feedback is a lot easier for some items on this list than others. If you are assessing a learner as a proceduralist, you can assess their knowledge and directly observe their physical skills. For knowledge, there are obvious questions: do they know when to perform the procedure, when not to perform the procedure, and theoretically how to perform the procedure? That is a good starting point, and it is relatively easy. I can talk to my students about intubation and understand whether they understand the procedure. 

Many people find it harder to assess the practice side, but that is probably because most of us haven’t broken down procedures into their smallest components. (If you just watch someone take a full golf swing, it will be hard to give feedback. But if you focus specifically on their grip, or on their take away, you can more easily provide feedback about that specific aspect of the swing.) If, like Scott Weingart has long been preaching, you focus on the microskills of our procedures, I think you will find it much easier to assess learners, and therefore give valuable feedback. 

When watching an intubation, I try to separate my assessment of visualization from that of tube delivery. Then, I try to break those down into their components. Was the patient positioned appropriately? Was the laryngoscope midline? Did they find the uvula; the epiglottis? Did they seat the laryngoscope in the vallecula? If they were still struggling, did they lift the neck, or manipulate the larynx with their other hand? By focusing on these minute details, I can provide excellent feedback on the procedure. If I just watch the procedure as a whole, I might know that it failed, but my feedback will be unhelpful. 

However,I haven’t broken down most aspects of medicine to the degree that I have airway management. When I watch a student perform a history, or a physical exam, I often struggle to give feedback. It is probably because I don’t know exactly what I am watching for. 

When watching a learner’s communication with patients, I know immediately when I think it went well and when it went poorly. However, I often struggle to provide feedback, because my assessment was based on gestalt rather than specific observations. To improve, can I break communication down into specific microskills, like I do procedures?

Some aspects might be obvious. Did the learner use jargon or other complex language? Did they check with the patient to ensure there was understanding? What was the speed of their speech? What was the tone? I am struggling here, because I have not seen this done before, so I am hoping that readers can help flush this out. 

Like communication, there are many aspects of medicine that I have historically assessed ‘holistically’, which might be impairing my ability to provide good feedback. For example, when assessing a learner’s differential diagnosis, I can state whether I would have come up with the same list, but after that relatively superficial assessment, I get somewhat stuck. It is hard for me to go deeper because I am not sure what I am observing. What was their process? How did they develop the differential? How did they move items up or down the list? How do I give great feedback if I am not even clear on the process I want them to follow?

What are the core skills I need to be observing in order to assess learners and provide feedback? Perhaps more importantly, how do I break those skills down into smaller microskills so I can provide specific, objective, and valuable advice? 

What about skills that take time to assess? How do we assess curiosity and drive in an emergency medicine rotation? If I get the opportunity to work with the same learner multiple times, I usually get a holistic sense of whether they are improving, but what do I say if they are not? What microskills should I be watching for?  What is my checklist to determine whether a student is a great lifelong learner?

Summary

I clearly have more questions than answers. I am convinced of one point: it is impossible to provide good feedback if you have not first defined great. We need a shared paradigm of a great physician. So please, comment below, what are the characteristics of the ideal doctor? And for those characteristics, what specific behaviours can we assess and discuss as part of the high quality feedback our learners deserve?

Other resources on feedback in medicine

EMCases WTBS 11 – Keeping Score: Providing Physician Feedback

STIMULUS podcast: Feedback can be hard to give and harder to receive. Here are techniques to do both better

The University of Toronto has a nice curated list of papers on feedback here

CanadiEM: “Help us help you” – Soliciting Feedback in the ED

ALiEM: Feedback in the Emergency Department