A prospective cohort study is conducted to evaluate the association between long-term proton pump inhibitor (PPI) therapy and the incidence of chronic kidney disease (CKD) in adults over age 50. After 5 years of follow-up, the investigators calculate the unadjusted (crude) relative risk of CKD. Suspecting that baseline diabetes mellitus might influence the outcome, they subsequently stratify the data by diabetes status. The results are shown in the table below:
| Patient Group | Relative Risk (RR) | 95% Confidence Interval |
|---|---|---|
| Crude (Unadjusted) | 1.9 | 1.4 - 2.5 |
| Stratum 1: Diabetes | 3.5 | 2.6 - 4.7 |
| Stratum 2: No Diabetes | 1.1 | 0.8 - 1.5 |
Based on these statistical findings, which of the following is the most appropriate next step for the investigators when reporting their results?
- Report the stratified relative risks separately for patients with and without diabetesAnswer
- BUse the Mantel-Haenszel method to calculate a single adjusted relative risk across both strata
- CPerform multivariable regression to adjust for diabetes as a confounding variable
- DConduct a new study utilizing strict matching for baseline diabetes status
- ERestrict the final analysis strictly to the cohort without diabetes to prevent selection bias
Answer
Report the stratified relative risks separately for patients with and without diabetes
The data demonstrates effect modification (also known as effect measure modification). This phenomenon occurs when the magnitude of an association between an exposure (PPI use) and an outcome (CKD) differs depending on the level of a third variable (diabetes). In effect modification, the stratified relative risks are significantly different from one another (e.g., 3.5 vs. 1.1). Because this represents a true biological interaction rather than a flaw in the study design, the most appropriate statistical approach is to report the stratified results separately rather than calculating an adjusted overall measure.
Step-by-Step Solution
Key Concept
Effect modification (effect measure modification) occurs when the magnitude of an effect differs based on a third variable. It is a true biological interaction that should be reported with separate stratified results, unlike confounding which should be adjusted.