A prospective cohort study is conducted to evaluate the association between chronic proton… — USMLE Step 3

A prospective cohort study is conducted to evaluate the association between chronic proton pump inhibitor (PPI) use and the incidence of gastric adenocarcinoma over a 15-year period.

Initial analysis reveals a crude relative risk (RR) of $2.8$ (95% CI: $2.1$ - $3.7$ ) for gastric adenocarcinoma in chronic PPI users compared to non-users. Concerned about the potential influence of Helicobacter pylori infection, the investigators perform a stratified analysis.

Among patients with confirmed H. pylori infection, the RR for developing gastric adenocarcinoma with chronic PPI use is $1.1$ (95% CI: $0.8$ - $1.5$ ). Among patients without H. pylori infection, the RR is $1.2$ (95% CI: $0.9$ - $1.6$ ).

Based on these stratified results, which of the following is the most appropriate approach for the investigators regarding the final analysis and reporting of this data?

Calculate and report an adjusted summary relative risk to account for the identified third variable.Answer
B
Report the stratum-specific relative risks separately to demonstrate that the variable modifies the biological effect.
C
Report the crude relative risk of $2.8$ , as the variable acts as a downstream mediator on the causal pathway.
D
Compare the mean incidence of adenocarcinoma between the two strata using a Student's two-sample t-test.
E
Calculate the positive predictive value of PPI use for gastric adenocarcinoma to determine the true clinical risk.

Answer

The most appropriate approach is to calculate and report an adjusted summary relative risk to account for the identified third variable.

The data provided in the vignette classically describes confounding. A confounder is an extraneous variable that correlates with both the dependent variable (gastric cancer) and the independent variable (PPI use). The hallmark of confounding during a stratified analysis is that the stratum-specific relative risks are similar to each other, but both differ significantly from the crude relative risk. In this scenario, the initial crude RR of

2.8

was spuriously elevated because H. pylori causes both dyspepsia (leading to PPI use) and gastric cancer. When stratified, the true RRs within each group dropped to approximately

1.0

. When confounding is identified, the correct epidemiological procedure is to calculate an adjusted summary estimate (using methods such as Mantel-Haenszel or multivariable regression) to provide a single risk estimate that controls for the confounding variable.

Step-by-Step Solution

Analyze the crude relative risk provided in the study.

The crude RR is

2.8

, suggesting a strong, statistically significant association between PPI use and gastric adenocarcinoma.

Establishing the baseline (unadjusted) association is necessary before evaluating the impact of stratification.

Evaluate the relative risks after stratification by H. pylori status.

The stratum-specific RRs are

1.1

and

1.2

. Both are close to

1.0

(null) and their 95% confidence intervals cross

1.0

Comparing stratum-specific measures to the crude measure differentiates between confounding, effect modification, and lack of third-variable influence.

Determine the epidemiological phenomenon occurring based on the data pattern.

Because the stratum-specific RRs are similar to each other (

1.1

and

1.2

) but markedly different from the crude RR (

2.8

), H. pylori is acting as a confounder.

Confounding distorts the crude association. Once the data is stratified by the confounder, the true (in this case, null) association is revealed within each stratum.

Select the correct statistical handling and reporting method for this phenomenon.

The investigators must calculate an adjusted summary relative risk (e.g., Mantel-Haenszel) to control for the confounding variable.

Unlike effect modification where strata are reported separately, confounding must be analytically controlled so a single, accurate, adjusted risk estimate can be reported.

Key Concept

Confounding versus Effect Modification

Question

Answer

Step-by-Step Solution

Key Concept

Common Mistakes

Topics

More questions from this topic