Confounding by indication — MRCPsych Paper A MCQ

Instant feedback + full explanation. One question, done properly.

ModerateEpidemiology and research methodsConfounding by indicationMRCPsych Paper A

An observational study finds higher mortality among people prescribed antipsychotics than among those not prescribed them. More severe illness both increased the likelihood of receiving an antipsychotic and was independently associated with mortality. What is the most likely explanation for the observed association?

Educational content. Not a substitute for clinical judgement or local policy.

Reveal the answer and explanation

Correct answer: A — Confounding by indication

The correct answer is A, confounding by indication. This occurs in an observational treatment study when a factor that influences prescribing also affects the outcome. Here, greater illness severity makes antipsychotic prescribing more likely and independently increases mortality risk, potentially producing a non-causal association between treatment and death. Regression to the mean describes unusually extreme measurements tending to become less extreme on repetition. Verification bias occurs when confirmation with a reference standard depends on an earlier test result. Lead-time bias creates an apparent survival advantage through earlier diagnosis without changing the time of death. Observer expectancy bias arises when an observer's expectations systematically influence measurement or interpretation. None explains the shared relationship of illness severity with both prescribing and mortality.

Reference: National Institute for Health and Care Excellence. NICE real-world evidence framework, section 'Introduction to real-world evidence in NICE decision making: Confounding'. Published 2022; updated 2025. https://www.nice.org.uk/corporate/ecd9/chapter/introduction-to-real-world-evidence-in-nice-decision-making