Genetics of Disease (Medical Interventions)
Unit 1: Clinical Reasoning LadderFeeds no tested domain

Clinical Reasoning 7: Order the test that separates

A test earns its place by splitting the differential: rival candidates should predict different results. Match the tool to the question, then break ties on risk.

Builds on (2 levels back)manual · high confidence
  • Clinical Reasoning 6: Build the differential: Rung 6 of the clinical ladder: a test is chosen to split a list, so the list must exist before the order does.

Re-learn the skill with worked practice and clear examples.

The real skill is the split: before ordering, write what each candidate predicts the result will be. If the predictions differ, the test separates. If they match, it cannot.

Step 1: Write the predictions first
Take the top two candidates and ask what each one expects this test to show. Two different expectations means the result will carry information.
Step 2: Spot the empty order
A test every candidate expects to pass, or every candidate expects to fail, returns a result you could have written in advance. It spends time and money to move nothing.
Step 3: Aim at the biggest fork
When several tests split the list, prefer the one whose branches change the plan most.
Practice

Two candidates remain for a teenager with a lingering cough: an asthma flare and early pneumonia. Which order does the most useful work next?

Reviewed
  1. A.A chest X-ray, because a pneumonia patch can show on the film while a flare usually leaves it clear
  2. B.A second telling of the history, since stories tend to drift on the second pass
  3. C.A height and weight measurement, to anchor the growth chart before treatment begins
  4. D.An hourly temperature log kept through the whole day, since either candidate can move a fever in either direction
Show the worked solution ▾

Answer: A. A chest X-ray, because a pneumonia patch can show on the film while a flare usually leaves it clear

  1. Step 1: Write both predictions: Pneumonia predicts a visible patch on the film. A flare predicts a clear film. The predictions differ, so the result will separate them.
  2. Step 2: Audit the rivals for splits: A fever log moves under both candidates, growth numbers move under neither, and a retold history adds noise rather than a fork.

Why it's right: The film is the one order whose result differs by candidate: a patch supports pneumonia and a clear film leans flare, so the result must move the list whichever way it lands.

Why the others miss:
  • B: A re-interview can be right or drifted; either way it is not a result the two candidates predict differently.
  • C: Baseline measurements have their uses, but both candidates predict the same chart, so no split.
  • D: Both candidates can run fevers, so the log's result reads the same under either and separates nothing.

Aligned to Clinical reasoning: the discriminating order · reading level ~grade 9

Where you'd see this
  • Utilization reviews ask one question of every imaging order: what were the rival predictions? No fork, no approval.
Guided notes

Fill these in as you work through the lesson.

Big idea: Order the investigation whose result will differ depending on which candidate is true; a test both candidates would pass, or both would fail, separates nothing and costs the patient anyway.
Key terms: write the meaning
  • Discriminating test (rival candidates predict different results on it):  
  • Modality (which tool: film, sound, magnet, tracer):  
  • Radiation cost (the tiebreaker between equal answers):  
  • Refine (what a good result does to the list):  
The rule

Before ordering, write each candidate's predicted  . If the predictions match, the test separates  . When two tests answer equally, the one with less   goes first.

Check yourself
  1. Why is a test both candidates would pass a waste even when it is cheap? 
  2. Which imaging tool sees bone best, and which sees stones without radiation? 
  3. What should be written down before a test is ordered, so the result can actually be scored? 
Work one example

Your differential for knee pain after a fall is a ligament tear versus a cracked kneecap. Write what an X-ray would show under each candidate, decide whether it separates them, and say what you would order if it does not.