Biostatistics and Research Methodology
Biostatistics and Research Methodology tests study design, data interpretation and critical appraisal. The exam asks whether a study’s design and analysis support the clinical claim being made.
Core rules
- Randomized controlled trials best test efficacy when randomization, allocation concealment and blinding are adequate. Cohort studies measure incidence and natural history; case-control studies are efficient for rare outcomes but risk selection bias.
- Bias includes selection, information, recall and publication bias. Confounding occurs when a third variable distorts the exposure–outcome relationship.
- A p-value is the probability of data this extreme if the null hypothesis is true; it is not the probability that the null is true. Confidence intervals show precision and clinical plausibility.
- Intention-to-treat analysis preserves randomization; per-protocol analysis can exaggerate benefit.
Worked example
In a trial, the control event rate is 20% and the treatment event rate is 12%. Absolute risk reduction = 20% − 12% = 8%, or 0.08. Number needed to treat = 1/0.08 = 12.5, rounded to 13. This means about 13 patients must be treated for one additional patient to benefit over the study period.
Traps
- Saying p < 0.05 proves a large or important effect.
- Ignoring confidence intervals and relying only on p-values.
- Confusing statistical significance with clinical significance.
- Preferring per-protocol results over intention-to-treat without justification.