RANZCP Critical Analysis Problem (CAP): The Statistics You Must Know

The Critical Analysis Problem trips up clinically strong candidates because it is a different skill: reading a paper, judging its design, and interpreting its numbers. The good news is that the same handful of concepts come up again and again.

The core concepts that recur

• Study designs and their hierarchy — what a cohort study can claim that a case-control study cannot.
• Diagnostic test metrics — sensitivity, specificity, and how positive predictive value depends on prevalence.
• Effect measures — relative risk vs absolute risk reduction, and the Number Needed to Treat (NNT = 1 / absolute risk reduction).
• Confidence intervals and p-values — what "statistically significant" does and does not mean.
• Bias and confounding — the usual suspects and how good design controls them.

Appraise a paper the way the exam wants

Work in a fixed order: the question and design first, then the methods (population, allocation, blinding, follow-up), then the results (effect size and precision), then the threats to validity, and finally what it means for practice. A consistent order means you never freeze in front of an unfamiliar paper.

Fellowship Ready includes a plain-language CAP statistics course — every concept above, with worked examples and diagrams, each section with a quiz. The overview and first section are free to read. Have a look — it is the fastest way to make the CAP feel routine.