Clinical Reasoning 2017

This is the first draft of a curriculum I’m writing to help pre-clinical students (starting with M1 year) start the process of clinical reasoning. I’m trying to go stepwise so it’s not too overwhelming, but I’m not sure if I’ve made it too simplified. I’d really love some feedback from students on what you think. The newer version can be found here.

Framework

Clinical Reasoning Algorithm

This is the framework we’ll be using throughout these videos.

01: Introduction

This video introduces the concept of clinical reasoning, as a brief overview of all the components. I haven’t recorded it yet because I wanted to finish all the other videos and then use those to make this one.

02: Illness Scripts

Illness scripts are the descriptions of a disease and contain several components:

  • Pathophysiology – what causes a disease?
  • Epidemiology – who gets the disease?
  • Timing – when it starts, when it stops, and what happens in between?
  • Symptoms and signs – What might you find on history and physical exam?
  • Diagnostics – what testing would you need to help rul-in or rule-out this disease?
  • Treatment – how would you treat the disease once ruled-in?

It may be useful to structure your pre-clerkship studies by arranging what you learn about diseases in illness scripts or even better illness script tables which compare 3 or more diseases. This is how you’ll be using this medical knowledge clinically in the future.

03: Problem Representation

This is how you structure the data you gather from your patients. You will learn the skills on how to do this in other classes (history taking, physical diagnosis, interpreting labs and x-rays, etc). Here we talk about what you do with that information. If you structure it properly going from problem lists, processing it using medical terminology including semantic qualifiers and finally structuring it into a patient illness script (problem representation), you’ll be able to use this to compare the patient presentation to various diseases in your differential diagnosis later.

04: Make a Differential Diagnosis

The differential diagnosis (DDx) is one of the most important parts of clinical reasoning. There’s an adage in medicine, “if you don’t think of it, you’ll never find it.” That means, never rely on blind luck or shotgunning (gathering a ton of non-specific information) giving you the answer. You need to know what you are looking for in order to find it.

So formation of a differential diagnosis is an important skill. There are several ways to do it.

  • MOST PROBABLE: think of what the most common causes of the patient’s symptoms are
  • MUST NOT MISS: what are dangerous causes of symptoms, though they may not be as common, that if missed can significantly harm the patient
  • ANATOMICAL: given where the patient’s symptoms are, what organs are in that area which could be the culprit?
  • PHYSIOLOGICAL: given how diseases work, what disease processes could cause the patient’s symptoms
  • MNEMONICS: there are certain memory aides that can help you create a complete differential diagnosis: AEIOU TIPS (for confused patients) or VINDICATE (to consider all possible organ systems)
  • PATTERN RECOGNITION: this is what expert clinicians do. They can recognize a disease intuitively without even thinking by drawing upon their years of experience. You will not start here.

05: Probability and Thresholds

We discuss probability here. For more information, see the corresponding videos in the EBM section. After collecting data, you assign each diagnosis in your differential a probability of being correct. While probability ranges from 0% to 100%, in reality we can never have certainty in medicine so even if you’re pretty darn sure, you’ll never achieve 0% or 100% probability.

The goal is to push the probability of a disease above two thresholds where we are “sure enough” about a disease to act on it (either treat it or discard/trash it). The in-between zone (test it), requires that you search for more information.

06: Sensitivity, Specificity & Predictive Values

These are the characteristics of tests. You can also find more information on how to calculate these within the EBM section.

07: Estimating Pre-Test Probability

Here we talk about how to estimate the initial probability of a disease before we order tests. Again our goal is to push the probability up above the treatment threshold or down below the testing threshold.

08: Introduction to Bayesian Reasoning

Bayesian reasoning describes the process of using testing to push a probability up or down using testing. Positive tests should push the probability of a disease up and negative tests push it down. The amount that the probability is pushed up or down depends on the quality of the test. Those qualities are described by the sensitivity and specificity of a test.