MRCOG Part 1: Statistics & Epidemiology
Statistics and Epidemiology is one of the most formula-driven subjects in MRCOG Part 1, with approximately 75 questions. The same question types appear repeatedly — sensitivity/specificity, positive predictive value, study design and statistical tests. Learn the formulas once and this becomes a high-scoring subject.
Topics Covered
Screening & Diagnostic Tests
- ›Sensitivity — proportion of true positives correctly identified (TP / TP + FN)
- ›Specificity — proportion of true negatives correctly identified (TN / TN + FP)
- ›Positive predictive value (PPV) — TP / TP + FP; depends on prevalence
- ›Negative predictive value (NPV) — TN / TN + FN; depends on prevalence
- ›Likelihood ratios — positive LR = sensitivity / (1 - specificity)
- ›ROC curves — area under curve measures test performance; top-left = ideal test
- ›Effect of prevalence on PPV and NPV — high prevalence increases PPV
Study Design
- ›Randomised controlled trial (RCT) — gold standard for causation, blinding types
- ›Cohort study — prospective or retrospective, relative risk calculated
- ›Case-control study — odds ratio calculated, good for rare outcomes
- ›Cross-sectional study — prevalence, not incidence
- ›Systematic review and meta-analysis — forest plots, heterogeneity (I²)
- ›Bias types — selection bias, information bias, confounding
Statistical Tests & Measures
- ›Mean, median, mode — which to use in skewed distributions
- ›Normal distribution — 68-95-99.7 rule
- ›Standard deviation vs standard error — SE = SD / √n
- ›p-value — probability of result if null hypothesis were true; p<0.05 convention
- ›Confidence intervals — 95% CI; if CI for RR includes 1.0, result is non-significant
- ›Type I error (false positive) vs Type II error (false negative); power = 1 - Type II error
- ›Parametric vs non-parametric tests — t-test, ANOVA, Mann-Whitney, chi-squared
Measures of Association
- ›Relative risk (RR) — cohort studies; RR = 1 means no association
- ›Odds ratio (OR) — case-control studies; approximates RR when disease is rare
- ›Number needed to treat (NNT) = 1 / absolute risk reduction
- ›Absolute risk reduction (ARR) vs relative risk reduction (RRR)
- ›Attributable risk — excess risk due to exposure in exposed group
Exam Tips for Statistics & Epidemiology
Sensitivity/specificity calculations from 2×2 tables are tested every sitting — practice until automatic.
Remember: PPV and NPV depend on prevalence; sensitivity and specificity do not.
Confidence intervals: a 95% CI for an odds ratio that crosses 1.0 means no significant association at p=0.05.
Study design question tip: if the question asks which study design is "most appropriate" for a rare disease, answer case-control.
Statistics at Square One (Swinscow) is free online — read it twice. It covers everything you need.
Recommended Book
Practice Statistics & Epidemiology questions now
AceMRCOG has 75 Statistics & Epidemiology SBA questions with detailed explanations. Start with 50 questions free — no credit card needed.
Start Free →