Pretest Probability

Pretest probability is the estimated likelihood of a condition being present before a diagnostic test result is known. It is one of the most important concepts in clinical reasoning — and one of the most frequently violated in informal medical education.

The concept comes from Bayesian reasoning. A test result does not tell you whether a patient has a disease. It tells you how much to update your estimate of the probability that the patient has the disease, given what you already knew before the test. A positive troponin in a patient with crushing chest pain, diaphoresis, and ST-elevation on ECG means something very different from a positive troponin in a young, healthy patient with atypical symptoms and a normal ECG. The test is the same; the pretest probability is different; the clinical meaning is different.

In FOAM contexts, pretest probability serves as a cognitive anchor against two common reasoning errors:

• Availability bias — overestimating the likelihood of dramatic or memorable diagnoses because FOAM content disproportionately features unusual cases. A podcast episode on a rare cause of chest pain can shift a listener’s mental model such that they start considering that diagnosis more often than its actual prevalence warrants. Pretest probability reminds the clinician to ask: how likely is this diagnosis in my patient population, regardless of how vivid the teaching case was?

• Base-rate neglect — ignoring the frequency of a condition in the relevant population when interpreting a test result. FOAM content sometimes emphasizes the sensitivity or specificity of a test without discussing the population in which the test is being applied. A highly sensitive test applied to a low-prevalence population generates many false positives — a fact that pretest probability makes explicit.

Pretest probability is not a number that clinicians calculate precisely in most clinical encounters. It is a disciplined habit of thought: before ordering a test, before interpreting a result, before acting on a FOAMed recommendation — ask how likely this diagnosis was before the new information arrived. That estimate shapes what the new information means.

Related terms

• Critical Appraisal — the broader discipline of evaluating evidence quality
• FOAM — the context in which pretest probability anchors interpretation