Explaining Your Clinical Reasoning: How to Answer 'Why?' on Rounds

Maya Brooks

Hi everyone, and welcome back to the AI Med Tutor podcast. I’m Maya Brooks, your AI-generated fourth-year medical student co-host. I bring the student perspective to our conversations, including the challenges we encounter, the questions we ask, and the things that can make medical school and board preparation harder than they need to be. Dr. Clinch, I want to talk about one word that can make even a well-prepared medical student suddenly forget everything they’ve learned. That word is, “Why?” Specifically, when you’re presenting a patient, you offer your assessment, and your attending looks at you and asks, “Why do you think that?” Sometimes I know the diagnosis is probably correct, but I struggle to explain the evidence. Other times, I feel pretty confident until the attending asks one question that makes me realize I’ve overlooked something important. I want to get better at both situations.

Dr. Randy Clinch

Hi everyone. I’m Dr. Randy Clinch, a DO family medicine physician and medical educator. Maya, that’s an excellent topic because it gets at something fundamental about clinical reasoning. Recognizing a diagnosis is one skill. Explaining why that diagnosis fits the evidence is another. And being willing to reconsider your assessment when new information emerges is another skill still. Today, I want us to explore how students can make their reasoning visible, respond thoughtfully when challenged, and use those moments to develop stronger clinical judgment. We’ll work through two clinical encounters, one in which your initial assessment is reasonable and another in which it needs to change. And before we go further, this episode is for education, not medical advice, and nothing we discuss is sponsored by any resource or vendor.

Maya Brooks

Let’s start with the first patient. I’m on an internal medicine rotation, and I’ve just evaluated a 42-year-old woman who came to clinic because of fatigue and shortness of breath with exertion. She says these symptoms have gradually worsened over the past several months. She also reports heavier menstrual bleeding over the past year. Her vital signs are stable, and she doesn’t appear acutely ill. Her complete blood count shows a hemoglobin of 9.8, an MCV of 70, and an elevated red cell distribution width. Her ferritin is 8. When I present her to my attending, I say that iron deficiency anemia is my leading diagnosis.

Dr. Randy Clinch

And your attending asks, “Why iron deficiency anemia?”

Maya Brooks

Right. And my first response might be, “Because she has a microcytic anemia and heavy menstrual bleeding.” Which isn’t wrong, but it feels incomplete.

Dr. Randy Clinch

It is a reasonable beginning. But let’s examine what your attending may be asking. They probably aren’t simply asking you to repeat the diagnosis. They want to know what evidence supports your conclusion, how strongly it supports it, and whether you’ve considered competing explanations. So what additional information strengthens your assessment?

Maya Brooks

The low ferritin is particularly important because it supports depleted iron stores. The low MCV indicates microcytosis, and the elevated red cell distribution width is consistent with variation in red blood cell size that can occur with iron deficiency. Her history of heavy menstrual bleeding provides a plausible source of ongoing iron loss.

Dr. Randy Clinch

Good. Now notice that those findings don’t all carry equal diagnostic weight. Microcytosis tells you something about the type of anemia, but several conditions can cause microcytosis. The low ferritin is much more specific evidence of iron deficiency in this clinical context. And the bleeding history helps explain why the deficiency may have developed.

Maya Brooks

So rather than just listing every abnormal finding, I should emphasize the evidence that actually distinguishes my diagnosis from the alternatives.

Dr. Randy Clinch

Exactly. Now imagine the attending asks, “Why not thalassemia trait?”

Maya Brooks

I’d say that thalassemia trait can also produce microcytosis, so the low MCV alone doesn’t distinguish the two. But the markedly low ferritin supports iron deficiency, and the elevated red cell distribution width and history of heavy menstrual bleeding fit that diagnosis. I wouldn’t claim thalassemia trait is impossible, but iron deficiency is much better supported by the information we have.

Dr. Randy Clinch

That’s a much stronger explanation. You’re identifying the leading diagnosis, supporting it with the most meaningful evidence, and showing that you’ve considered an alternative. You’re also avoiding an unnecessary claim of certainty.

Maya Brooks

I think that’s where students sometimes get stuck. We feel as though we need to sound completely certain when presenting a patient. So we say, “This patient has iron deficiency anemia,” as though there’s no uncertainty. But when the attending asks us to defend that conclusion, we may discover that we haven’t thought through the evidence carefully enough.

Dr. Randy Clinch

And there’s a difference between communicating a clear assessment and pretending to have absolute certainty. You can say, “Iron deficiency anemia is my leading diagnosis because of the low ferritin, microcytosis, and history of heavy menstrual bleeding.” That’s clear and appropriately confident. You don’t need to weaken every sentence with multiple disclaimers, but you should recognize what the evidence does and doesn’t establish.

Maya Brooks

There’s another piece of this case that I think matters. Even if I’m correct that she has iron deficiency anemia, I haven’t finished my clinical assessment. I still need to consider the cause of her iron deficiency, whether the menstrual bleeding adequately explains it, and what further evaluation is appropriate.

Dr. Randy Clinch

That’s important. Identifying iron deficiency doesn’t automatically establish the source of blood loss. You’d want to consider the patient’s history, risk factors, and other potential causes, including gastrointestinal blood loss when indicated. Your attending may challenge you not because your diagnosis is wrong, but because your reasoning stops too early.

Maya Brooks

So the question “Why?” might mean several things. Why is this the leading diagnosis? Why is it more likely than another diagnosis? Why does this finding matter? Or why are you recommending a particular next step?

Dr. Randy Clinch

Right. And you can use the question to clarify the reasoning task. If your attending asks why you want a particular test, explain what clinical uncertainty that test is intended to resolve and how the result might affect your decision. If they ask why you favor one diagnosis, explain the evidence that distinguishes it from the alternatives.

Maya Brooks

Let’s try that. Suppose I tell my attending that I want to evaluate the cause of her heavy menstrual bleeding and consider whether other sources of blood loss need investigation. The attending asks why. I’d explain that confirming iron deficiency identifies the problem, but understanding the source is necessary to address the cause and reduce the likelihood of recurrence.

Dr. Randy Clinch

Good. You’ve connected the proposed evaluation to a clinical purpose. That’s the kind of reasoning that turns a list of tests or treatments into an actual assessment and plan.

Maya Brooks

I can see how this could improve my patient presentations. Instead of memorizing a long script, I could practice giving a concise conclusion and then explaining the evidence that supports it. Something like, “My leading diagnosis is iron deficiency anemia. The strongest supporting findings are the low ferritin, microcytosis, and history of heavy menstrual bleeding. Thalassemia trait is another consideration, but the available evidence favors iron deficiency.”

Dr. Randy Clinch

That’s a useful structure. And it doesn’t need to be lengthy. A strong clinical explanation often takes only a few sentences. The goal isn’t to overwhelm your attending with everything you know. It’s to show how you reached your conclusion.

Maya Brooks

I also want to talk about what happens when I don’t know the answer. Suppose the attending asks me something about the interpretation of the red blood cell indices, and I realize I can’t explain it confidently. What should I do?

Dr. Randy Clinch

Be honest about what you know and what you don’t. You might say, “I know that the elevated red cell distribution width can support iron deficiency in this context, but I’m not confident I can explain the full mechanism. I’d like to review that.” That’s more useful than inventing an explanation or repeating the same answer with greater confidence.

Maya Brooks

And if I do review it later, I should be able to explain the mechanism rather than just memorize the phrase that elevated RDW goes with iron deficiency.

Dr. Randy Clinch

Right. That would be a meaningful learning repair. You’re connecting the laboratory finding to the underlying process so you can use that knowledge again in a different patient.

Maya Brooks

Okay. Now I want to move to the second patient because this one is more uncomfortable. In the first case, I had the right leading diagnosis and needed to explain it better. In this next case, I’m going to make an assessment that sounds reasonable at first, but it turns out I’ve overlooked something important.

Dr. Randy Clinch

Let’s hear the case.

Maya Brooks

I’m on an emergency medicine rotation, and I’m evaluating a 56-year-old man who comes in with right-sided upper abdominal pain, nausea, and vomiting. He says the pain began after dinner and has persisted for several hours. He has tenderness in the right upper quadrant, and his temperature is 38.5 degrees Celsius. His heart rate is 108, blood pressure is 124 over 76, and oxygen saturation is normal. I know that gallstones can cause pain after meals, so when I present the patient, I say that biliary colic is my leading diagnosis.

Dr. Randy Clinch

And the attending asks, “Why biliary colic?”

Maya Brooks

I’d say that he has right upper quadrant pain after eating, which fits a gallstone-related process.

Dr. Randy Clinch

And what about the fever and tachycardia?

Maya Brooks

That’s where I start to realize my reasoning is incomplete. Those findings are concerning for an inflammatory or infectious process. Biliary colic usually doesn’t cause fever. So I may have recognized a familiar pattern but ignored findings that don’t fit it.

Dr. Randy Clinch

Good. Now let’s slow down and examine what happened. You recognized right upper quadrant pain after a meal and connected it to gallstones. That’s reasonable. But you anchored on that association before adequately considering the rest of the presentation. Persistent pain, fever, and tachycardia should prompt you to consider acute cholecystitis and other potentially serious causes.

Maya Brooks

So the attending’s question isn’t simply asking me to explain my answer. It’s revealing that my answer may not account for the most important findings.

Dr. Randy Clinch

Exactly. And that’s an opportunity to revise your assessment. What would you say now?

Maya Brooks

I’d say, “On reconsideration, acute cholecystitis is more concerning than uncomplicated biliary colic because the pain has persisted and the patient has fever and tachycardia. Those findings suggest an inflammatory process rather than transient obstruction alone. I would want to evaluate for acute cholecystitis and assess for other biliary complications or alternative diagnoses.”

Dr. Randy Clinch

That’s a stronger assessment. And notice what you’ve done. You haven’t simply replaced one diagnosis with another because the attending seemed dissatisfied. You’ve identified the evidence that makes your original conclusion less convincing and explained why your revised assessment better fits the presentation.

Maya Brooks

I think that distinction matters. Sometimes when an attending challenges us, we’re tempted to immediately change our answer because we assume the attending must be pointing toward something else. But I shouldn’t just guess what answer they want. I should reconsider the evidence.

Dr. Randy Clinch

That’s an important point. Clinical reasoning isn’t a game of reading the attending’s facial expression. The purpose of the challenge is to examine your thinking. Sometimes your original assessment will remain the best explanation. Other times you’ll discover a reason to change it. The goal is to let the evidence guide that decision.

Maya Brooks

Now let’s say the attending asks what I’d recommend next. I know I need to consider the patient’s clinical stability, but his blood pressure is normal. Does that mean I can treat this as a routine, nonurgent problem?

Dr. Randy Clinch

No. A normal blood pressure doesn’t eliminate the possibility of a serious illness. This patient has fever, tachycardia, and persistent abdominal pain, which warrant prompt assessment. You’d evaluate for systemic illness and potential complications while pursuing the diagnostic workup. The specific management would depend on the complete clinical picture, but suspected acute cholecystitis generally requires timely evaluation, appropriate imaging, and consideration of surgical management.

Maya Brooks

So I shouldn’t confuse having a normal blood pressure with having no urgent clinical concerns. I’d assess the patient carefully, consider the possibility of infection or other complications, and discuss the appropriate evaluation and treatment with my supervising clinician.

Dr. Randy Clinch

Right. And this connects to the stability-first approach we’ve discussed in previous episodes. The patient’s condition determines how urgently you need to act, even while you’re refining the diagnosis.

Maya Brooks

I’d want to review laboratory studies and obtain appropriate imaging, typically right upper quadrant ultrasound as an initial study when acute cholecystitis is suspected. And depending on the patient’s findings, the team would consider treatments such as intravenous fluids, analgesia, antibiotics when indicated, and surgical consultation.

Dr. Randy Clinch

Good. And remember that your role as a student is to develop and communicate a reasoned assessment and proposed plan under supervision. You don’t need to pretend you’re making all the decisions independently. What matters educationally is that you can explain why you’re recommending the next step.

Maya Brooks

I want to go back to the moment when I realized my initial diagnosis was wrong. I think students can feel embarrassed when that happens, especially in front of an attending or the rest of the team. How should we handle that moment?

Dr. Randy Clinch

First, recognize that revising your assessment in response to evidence is a clinical reasoning skill, not a failure of character. You’re learning to integrate information and make judgments. If you discover that your initial conclusion doesn’t fit the evidence, say so clearly. You might say, “I initially favored biliary colic because of the postprandial right upper quadrant pain, but the persistent pain and fever make acute cholecystitis more likely.” That’s a thoughtful response.

Maya Brooks

And it’s much better than trying to defend my original answer by ignoring the findings that don’t fit.

Dr. Randy Clinch

Yes. One of the habits we want to avoid is defending a diagnosis simply because we were the first person to propose it. Your assessment is a working explanation, not a position you have to protect. The evidence matters more than preserving your first answer.

Maya Brooks

That reminds me of something we’ve discussed in our episodes about clinical reasoning errors. Anchoring can occur when we latch onto an early clue and give it too much weight. In this case, I focused on the pain after eating and didn’t adequately account for the fever and persistent symptoms.

Dr. Randy Clinch

That’s a good connection. And one way to counter that tendency is to ask yourself what findings your diagnosis doesn’t explain. In our first case, iron deficiency anemia accounted for the laboratory findings and the patient’s symptoms reasonably well, although we still needed to investigate the cause. In the second case, uncomplicated biliary colic didn’t adequately explain the fever and systemic findings. That mismatch should have prompted reconsideration.

Maya Brooks

So if I’m presenting a patient and I feel confident about my diagnosis, I could pause and ask, “What’s the strongest evidence against my conclusion?” Or, “What important finding am I not explaining?”

Dr. Randy Clinch

Those are excellent questions. And you don’t need to spend ten minutes doing that for every patient. Sometimes a brief check for contradictory evidence is enough to prevent premature closure.

Maya Brooks

I think there’s another practical lesson here. In both cases, the attending asked essentially the same question. “Why?” But in the first case, the answer was to articulate the supporting evidence more clearly. In the second case, the answer was to reconsider the diagnosis because the evidence didn’t support my initial conclusion as well as I thought.

Dr. Randy Clinch

Exactly. The same question can serve different purposes. Sometimes it helps you communicate sound reasoning. Sometimes it exposes a knowledge gap. Sometimes it reveals a thinking error. And sometimes it prompts you to recognize that your assessment needs to change.

Maya Brooks

So how can students practice this skill outside of actual patient presentations? I don’t want to turn this into another complicated study system, but I do think it would help to practice explaining my reasoning before I’m standing in front of an attending.

Dr. Randy Clinch

You can practice it with the clinical questions you’re already answering. After you choose an answer, ask yourself whether you could explain why it’s the best choice using the strongest findings in the question. Then identify one plausible alternative and explain why it’s less likely. You don’t need to do this for every question, but it’s particularly useful when you were uncertain or when two choices seemed reasonable.

Maya Brooks

And I could practice saying the explanation aloud rather than just thinking it. That would help me get comfortable expressing the reasoning in a concise way.

Dr. Randy Clinch

Yes. Speaking your reasoning can reveal gaps that aren’t obvious when you’re simply recognizing information on a page. You may discover that you can identify the correct answer but can’t explain the mechanism or the distinguishing evidence. That’s useful feedback.

Maya Brooks

I could also practice with a classmate. One of us presents a short clinical case and proposes a diagnosis, and the other asks why. Then we switch roles. The goal wouldn’t be to stump each other with obscure facts. It would be to examine the reasoning.

Dr. Randy Clinch

That’s a good approach. And you can make it more useful by asking different kinds of follow-up questions. What evidence supports your conclusion? What alternative did you consider? What finding would make you change your mind? And why is your proposed next step appropriate for this patient?

Maya Brooks

I especially like that last question. Because sometimes I can explain why a diagnosis is likely but struggle to explain why a particular test or treatment is the next best step.

Dr. Randy Clinch

And that’s another level of clinical reasoning. A test should address a meaningful clinical question. A treatment should address the patient’s condition and current needs. When you can explain the purpose of the next step, you’re moving beyond memorizing algorithms toward understanding the decisions within them.

Maya Brooks

Let’s make this even more practical. Suppose I have thirty seconds to explain my assessment during rounds. I don’t have time for a lengthy differential diagnosis. What would a strong response sound like?

Dr. Randy Clinch

I’d aim for a clear leading conclusion, the two or three most meaningful supporting findings, and a brief explanation of the main alternative or concern that affects the decision. For our first patient, you might say that iron deficiency anemia is the leading diagnosis because of the low ferritin, microcytosis, and history of heavy menstrual bleeding. You’d acknowledge that the cause of the iron deficiency still needs evaluation. That’s concise, evidence-based, and clinically useful.

Maya Brooks

And for the second patient, I’d say that acute cholecystitis is a leading concern because of persistent right upper quadrant pain, fever, and tachycardia. I’d explain that uncomplicated biliary colic doesn’t adequately account for those systemic findings, and I’d recommend prompt evaluation for acute cholecystitis and potential complications.

Dr. Randy Clinch

Exactly. You’re not trying to recite everything you know about gallbladder disease. You’re communicating the reasoning that matters for this patient’s current clinical problem.

Maya Brooks

I think this also changes how I approach feedback. If my attending challenges my assessment, I don’t have to interpret that as criticism of my ability. I can treat it as information about where my reasoning needs to become clearer or more complete.

Dr. Randy Clinch

That’s a productive way to approach it. And if the feedback is unclear, you can ask a focused question. For example, “Was there a finding I didn’t weigh appropriately?” Or, “Which feature most strongly distinguishes these diagnoses?” That invites specific feedback you can use.

Maya Brooks

And afterward, I could make a very small learning note. Not a complete rewrite of the patient encounter, but perhaps one sentence about the reasoning gap. For the abdominal pain case, it might be, “Persistent right upper quadrant pain with fever should prompt concern for acute cholecystitis rather than uncomplicated biliary colic.”

Dr. Randy Clinch

That’s useful. And if you want to reinforce it, you could retrieve that distinction later or apply it to another case. But the important thing is that you’ve identified the reasoning lesson. You don’t need to create an elaborate record of every question your attending asks.

Maya Brooks

I also want to acknowledge that some students may feel uncomfortable being questioned in front of others. The attending’s tone and the learning environment matter. A respectful challenge can be helpful, but feeling publicly embarrassed isn’t the same thing as receiving effective teaching.

Dr. Randy Clinch

I agree. Good clinical teaching should challenge reasoning while maintaining respect for the learner. Students benefit when they can explain their thinking, acknowledge uncertainty, and revise an assessment without being humiliated. And when the environment is supportive, asking why becomes a shared effort to understand the patient rather than a test of whether the student can survive the questioning.

Maya Brooks

So as students, we can work on making our reasoning visible and responding thoughtfully. And as educators, we should use questions to understand how students are thinking, not simply to see whether they can recall the answer we have in mind.

Dr. Randy Clinch

That’s well said. The question “Why?” is most valuable when it opens a conversation about evidence, mechanisms, alternatives, and decisions.

Maya Brooks

Let’s bring our two patients together. In the first case, I correctly recognized iron deficiency anemia, but I needed to explain why the low ferritin, microcytosis, and bleeding history supported that diagnosis. In the second case, I initially favored biliary colic because of the pain after eating, but I hadn’t adequately accounted for the fever and persistent symptoms. When the attending challenged me, I had to reconsider the evidence and revise my assessment toward acute cholecystitis.

Dr. Randy Clinch

And those are two different successes in clinical reasoning. One is the ability to explain a well-supported conclusion. The other is the ability to recognize when a conclusion needs to change. Both are essential.

Maya Brooks

I think the biggest takeaway for me is that being asked why isn’t necessarily a sign that I’ve done something wrong. It’s an invitation to show how I’m thinking. And if I can’t explain my reasoning, or if the evidence doesn’t support my conclusion, that’s something I can learn from.

Dr. Randy Clinch

Exactly. And with practice, you can become more comfortable making your reasoning explicit. You’ll learn to identify the strongest evidence, consider alternatives, communicate uncertainty appropriately, and revise your assessment when new information warrants it.

Maya Brooks

So here’s the challenge I’d leave with listeners this week. The next time you present a patient or answer a clinical question, don’t stop at naming the diagnosis. Ask yourself why you believe it’s the best explanation. Identify the strongest supporting evidence, consider one reasonable alternative, and think about what finding might change your mind. Then practice explaining that reasoning aloud in a few clear sentences. And if your attending challenges your assessment, use the question to examine your thinking rather than simply defend your first answer. Knowing the diagnosis matters, but being able to explain why it fits the patient is what makes your clinical reasoning useful. Thanks for spending part of your week with us on the AI Med Tutor podcast. And remember, stay curious, and keep learning.