Biostatistics and epidemiology for USMLE Step 1

A study follows 20,000 people over 15 years, recording smoking status at baseline and subsequent incidence of bladder cancer. Which measure of association can be calculated directly from this design?

1. Prevalence ratio That belongs to a cross-sectional study, which measures existing cases at a point in time rather than new cases over time.
2. Relative risk correct Correct. This is a prospective cohort study, in which incidence is measured directly in exposed and unexposed groups, so risk in each group and therefore relative risk can be calculated.
3. Hazard ratio only, with no other measure available A hazard ratio can certainly be estimated from time-to-event data, but it is not the only measure available, and the question asks what the design permits directly.
4. Odds ratio only That is the constraint of a CASE-CONTROL design, where sampling is on outcome so incidence cannot be measured. An odds ratio can also be computed here, but it is not the limitation.
5. Attributable risk cannot be calculated from this design Wrong. Attributable risk is exactly what a cohort design supports, being the difference in incidence between exposed and unexposed.

The point: Study design determines the measure. Cohort: incidence, relative risk, attributable risk. Case-control: odds ratio only. Cross-sectional: prevalence. Randomized trial: causal inference. Rare disease means the odds ratio approximates the relative risk.

Source: USMLE Content Outline — biostatistics and epidemiology NBME / FSMB (USMLE programme) · tier 0, exam blueprint / regulator

A screening test for a disease with a prevalence of 1% has a sensitivity of 99% and a specificity of 90%. It is applied to 10,000 people. Approximately what is the positive predictive value?

1. About 90% Confuses PPV with specificity. PPV depends on prevalence; specificity does not.
2. About 99% Confuses PPV with sensitivity, and would mean almost every positive is a true case.
3. About 9% correct Correct. Of 100 with disease, 99 test positive. Of 9,900 without, 10% (990) test positive falsely. PPV = 99 / (99 + 990) ≈ 9%. In a low-prevalence population even an excellent test generates mostly false positives, the single most examined idea in USMLE biostatistics.
4. About 50% Would require the number of true and false positives to be similar, which needs a far higher prevalence or a much better specificity.
5. It cannot be calculated without the incidence Prevalence is given, which is exactly what is needed. Incidence is irrelevant to predictive value.

The point: Sensitivity and specificity are properties of the TEST. PPV and NPV depend on PREVALENCE. As prevalence falls, PPV falls and NPV rises. Confirm a positive screen with a highly specific test, the reason for the ELISA then Western blot sequence.

Source: USMLE Content Outline — biostatistics and epidemiology NBME / FSMB (USMLE programme) · tier 0, exam blueprint / regulator

A cohort study finds that coffee drinkers have a higher rate of lung cancer than non-drinkers. Coffee drinkers in this population are also far more likely to smoke. After adjustment for smoking status, the association with coffee disappears. Which phenomenon does this illustrate?

1. Confounding correct Correct. Smoking is associated with the exposure (coffee) and independently causes the outcome (lung cancer), and it is not on the causal pathway between them. Adjustment, stratification, matching or randomisation removes it.
2. Recall bias A differential error in remembering exposure, which affects case-control studies far more than a cohort design with baseline measurement.
3. Berkson bias Selection bias arising from using hospitalised controls, which does not apply to a community cohort.
4. Effect modification In effect modification the association differs ACROSS strata, real in one group, absent in another, and it is reported separately rather than adjusted away. Here the association vanished entirely.
5. Lead-time bias The apparent survival gain from diagnosing a disease earlier without changing its course. Relevant to screening studies, not to this one.

The point: Confounder test: associated with the exposure, independently causes the outcome, not an intermediate step. Fix it at design (randomisation, restriction, matching) or at analysis (stratification, multivariable adjustment). Effect modification is not a bias, report the strata.

Source: USMLE Content Outline — biostatistics and epidemiology NBME / FSMB (USMLE programme) · tier 0, exam blueprint / regulator