Clinical Problem Solving: A student and a teacher talk about lessons learned from an online course

Post Thumbnail
By Heather Murray, MD, and Eve Purdy, MD Candidate, 2015

For many medical students, the process involved in turning a presenting complaint into an appropriate and focused differential diagnosis seems like a big black box. For clinicians who do this many times every day, the process is unconscious, and it is hard to explain to medical student learners how to break it down. Both students and teachers sometimes struggle with how to transition early medical learners to competent diagnosticians.

black boxSo, when a clinician (Heather Murray) and a second year medical student (Eve Purdy) independently stumbled across the link to a Massive Open Online Course (MOOC) on Clinical Problem Solving offered through Coursera both of us jumped at the opportunity to learn more about diagnostic reasoning. Eve registered with the hope of shedding light on the type of problem solving that she might be faced with in clerkship, while Dr. Murray registered with the intention of improving her teaching around diagnostic reasoning for students.

Though it is difficult to summarize the six-week course in one blog post there were a few takeaways from the course that we will outline. These key points might help medical students improve clinical reasoning and the same tips might help teachers in clarifying the process for learners. Much of this approach to clinical reasoning comes from the NEJM article  “Educational Strategies to Promote Clinical Reasoning” by Judith Bowen (2006).

1. Organize the way you learn about diseases using Disease Illness Scripts

If you have a structured approach to the way you learn about diseases, then you will be more efficient at recalling that information and comparing diseases effectively. One way to organize information is into “Disease Illness Scripts”. This requires organizing information about the conditions into four broad categories.

Epidemiology Timing Clinical Presentation Pathophysiology
-who gets the disease?-what are the risk factors?

-making a mental picture of who you would expect to see with the disease can help

-over what time period does the condition present? 

hyperacutely: hours

acutely: days

sub-acutely:    days-months

chronic: months-years

acute on chronic

-a good way to think about this is where you would expect to see the patient (ER, vs walk-in vs family doctor)

-what are the symptoms? 

-physical signs?

*key features are signs and symptoms that are essential to the diagnosis

*differentiating signs and symptoms are those that make this disease different then diagnoses that present similarly

*excluding signs and symptoms are those that, if present, exclude the disease

-describe and understand the underlying disease mechanism

2. Organize the way you think about patients using Patient Illness Scripts

When thinking about patients try to frame their presentation using the same structure as the disease illness scripts.

Epidemiology Timing Clinical Presentation
What important risk factors does the patient have-age

-smoking

-relevant medical history

-presentation specific risk factors i.e. recent transcontinental  air travel in a patient with shortness of breath

How long has the patient had the symptoms, have they changed?  What symptoms and clinical signs does the patient have? 

-try to group as many as possible to shorten the list (e.g. group febrile, tachycardic and hypotensive as septic)

3. Compare disease illness scripts and patient illness scripts to generate a tiered differential diagnosis

Generate a differential diagnosis based on the chief complaint. You can compare your understanding about each disease on your differential with your patient using the illness scripts easily. Pay close attention to key features, differentiating features and excluding features. The closer a disease illness script is to the patient illness script the higher it should end up on your differential. Your final differential has three tiers:

Tier 1: Diseases that are those most likely belong here. The epidemiology, time course and clinical presentation are concordant with the patient illness script.

  • Tier 1e: Diseases on tier 1e are diagnoses that may be less likely than tier 1 but if missed will cause immediate and serious harm. These are dangerous diagnoses! The “e” in this tier stands for “emergency” and diseases on this list must be ruled out, even if they are less likely.

Tier 2: Diseases that have some similarities to the patient illness script but aren’t a perfect fit belong here. They are still possible but less likely than tier 1 diagnoses.

Tier 3: Diseases on your original list that do not fit the illness script. They may have excluding features or lack key features.

 4. Use your tiered differential to determine what tests to order

The tier that a possible diagnosis falls into will help you decide what tests to order to determine the final diagnosis. Think of each tier as a pretest probability.

Tier 1 diagnoses have a “high” pretest probability

  • No tests or few tests may be needed to convince you that a diagnosis in tier 1 is responsible for the patient’s presentation and similarly you would need very convincing information to take it off your list completely.
  • These and Tier 1e diagnoses should drive your initial investigations

Tier 1e diagnoses may have varying pretest probability

  • These diseases may or may not be likely but regardless tests with high sensitivity are needed to rule them out (remember “SnOUT”)

Tier 2 diagnoses have a “medium” pretest probability

  • Diseases on this tier are tricky. You really have to evaluate the sensitivity, specificity and information given from each test. You may need a few good tests get from a “medium” pretest probability to final diagnosis.

Tier 3 diagnoses have a “low” pretest probability

  • Even relatively good tests may not move diagnoses from Tier 3 up to tier 1. The positive result that you get might be due to chance. Investigating these diagnoses should be a last resort.

Gear box

These four tips won’t magically turn a medical student into an expert at clinical reasoning but they might serve to expose the way that experts think. They offer concrete ways for medical students to approach clinical reasoning and a common language for experts to discuss their approach with their learners.

For more information about MOOCs and why explicit discussion of clinical reasoning is important, see these links.

Leave a Reply

Post Timeline

What’s the price of professional autonomy?
Are we willing to pay it?
Published Mon, May 25, 2015

We all need people in our lives who are willing to tell us “how it is”. These are special folks, often spouses, relatives or lifelong friends, whose relationship with us is strong enough to allow candid, honest, unvarnished commentary about topics close to the heart. I found myself in a conversation with one such person recently who had the temerity … Continue reading

When you are yourself, I’m free to be myself
Published Mon, May 18, 2015

“When you are yourself, I’m free to be myself” The Reverend Bill Hendry spoke these words as a “first patient” at the First Patient Program’s 3rd annual Grand Finale on Wed. May 13. He was addressing the 100 students of the class of 2017 who had completed their 18 plus months of relationship with their first patient, whom they’d met … Continue reading

Milestones: A tribute to our tenacious 2015 graduating class.
Published Mon, May 11, 2015

“One hundred and eighty-five”. That was the answer to my question. The question, that I’d posed somewhat naively to our intrepid assessment coordinator, Amanda Consack, was “how many assessments have the 2015 class undertaken during medical school?” “Do you mean everything?” she asked. “Yes. Everything”, I answered, not wanting to sound wimpy. In her typical fashion, Amanda provided me not … Continue reading

Student attention in class: Whose responsibility?
Published Mon, May 4, 2015

I just received a posting from Faculty Focus with the engaging title:  Why can’t students just pay attention?  Dr. Chris Hakala, the author, gives a really good overview of the dilemma many of us face when teaching:  students are not engaged, are multi-tasking at best, and distracted at worst, and are not learning or retaining key concepts.  How much responsibility … Continue reading

Queen’s student wins 2015 Sandra Banner Award for Leadership
Published Tue, April 28, 2015

Queen’s Class of 2015 student Eve Purdy received the CaRMS Sandra Banner Award for Student Leadership at the CaRMS forum held in conjunction with the Canadian Conference on Medical Education (CCME) in Vancouver on April 26. Richard Reznick, Dean of the Queen’s Faculty of Health Sciences, presented the award on behalf of the award selection committee. “Eve has always challenged … Continue reading