Treatment Effects, Inference, Regression and Missing Data
Risk and benefit measures, NNT and NNH, standard error, hypothesis tests, confidence intervals, errors, power, correlation, regression, intention-to-treat and missing data.
How to prepare for MRCPsych Paper B
Integrate clinical formulation, safety and current UK practice with disciplined appraisal of design, bias, estimates and applicability.
Core concepts
Concept 1
Treatment effects can be expressed in absolute and relative terms; confidence intervals show compatible effect sizes and precision, while hypothesis tests quantify compatibility with a null model.
Exam cue: Calculate event risks first, then absolute difference, relative effect and number needed to treat or harm with direction clear.
Concept 2
Regression adjusts or models associations under assumptions, and intention-to-treat analysis preserves randomised comparison; missing data can still bias either approach.
Exam cue: Interpret the estimate, confidence interval, clinical importance, assumptions and missingness rather than a p-value alone.
Risk pitfalls and guardrails
Presenting a large relative reduction without the baseline risk or absolute effect.
Guardrail: Do not convert a group association or score into certain individual prediction or a stand-alone disposition decision.
Calling a non-significant result proof of no effect despite a wide interval and low power.
Guardrail: Do not equate statistical significance with importance or a non-significant result with proof of no effect.
Memory anchors
What is absolute risk reduction?
Control-group event risk minus intervention-group event risk for a beneficial outcome framing.
How is number needed to treat derived?
It is the reciprocal of the absolute risk reduction when risk is expressed as a proportion.
What does a confidence interval convey?
A range of effect values compatible with the data and model at the stated confidence level, indicating precision.
What is a type I error?
Rejecting a true null hypothesis according to the chosen decision rule.
Why use intention-to-treat?
It analyses participants in their randomised groups, preserving the benefits of allocation and estimating the effect of assignment.
Checkpoint rule
Do the check-up only after you can summarize each concept in one sentence and identify one dangerous pitfall from memory.
Knowledge Check (after reading)
Short check-up to confirm understanding of this module.
Check-up Questions
Relapse occurs in 30% of controls and 20% of treated participants. What is the absolute risk reduction?
Using control risk 30% and treatment risk 20%, what is the relative risk?
Answer all questions to submit.
Next step personalized recommendations
Continue learning
Move forward only after this module is stable.
What is Pass Harbor?
Completely free exam prep for 247 UK exams.
- Practice questions
- Flashcards
- Study guides
- Mock exams
- No registration
- No paywall
- Start instantly
“No more expensive exam prep. Quality study tools should be accessible to everyone.”