A clinical research team is analyzing data from a recent prospective cohort study investig… — USMLE Step 3

A clinical research team is analyzing data from a recent prospective cohort study investigating a novel oral spleen tyrosine kinase (Syk) inhibitor for the treatment of chronic spontaneous urticaria. The primary outcome is achieving complete symptom control at $12$ weeks. The crude analysis demonstrates that patients receiving the Syk inhibitor have a higher likelihood of achieving complete symptom control compared to those on standard high-dose antihistamines, yielding a relative risk (RR) of $1.75$ ( $95\% \text{ CI}, 1.35-2.20$ ).

Upon further data review, the lead epidemiologist stratifies the analysis based on the presence of baseline autoimmune thyroid disease. The stratified results are as follows:

Patients with autoimmune thyroid disease: $RR = 1.08$ ( $95\% \text{ CI}, 0.85-1.38$ )
Patients without autoimmune thyroid disease: $RR = 2.65$ ( $95\% \text{ CI}, 1.95-3.55$ )

Based on these statistical findings, which of the following is the most appropriate approach for interpreting and reporting this study's data?

Report the outcome estimates separately for each thyroid disease stratum, as the baseline condition modifies the drug's therapeutic magnitude.Answer
B
Report a single, adjusted summary estimate using the Mantel-Haenszel method to control for the distortion introduced by the baseline condition.
C
Retain and report the initial crude relative risk, because stratifying the data inappropriately reduces the statistical power of the primary endpoint.
D
Exclude patients with baseline autoimmune thyroid disease from the final analysis to eliminate the source of selection bias.
E
Apply an analysis of variance (ANOVA) test to compare the symptom resolution rates between the two strata and determine if the crude relative risk is valid.

Answer

Report the outcome estimates separately for each thyroid disease stratum, as the baseline condition modifies the drug's therapeutic magnitude.

The correct answer identifies that the study demonstrates effect modification (also known as statistical interaction). When a study's crude relative risk (

RR=1.75

) differs significantly from its stratified relative risks, one must look at the strata. In confounding, the RRs in both strata would be similar to each other (e.g., both around

1.10

). In this scenario, the stratum-specific RRs (

1.08

and

2.65

) are starkly different from one another. This indicates that the presence of autoimmune thyroid disease biologically changes how the Syk inhibitor works. Because effect modification is a true natural phenomenon rather than a study flaw, the most appropriate scientific step is to report the stratum-specific estimates separately to inform clinical practice accurately.

Step-by-Step Solution

Analyze the crude data provided in the scenario.

The initial unstratified relative risk is

1.75

, suggesting a significant benefit overall.

Establishing the baseline finding is necessary to compare against stratified results.

Evaluate the stratum-specific data.

The relative risk for patients with thyroid disease is

1.08

(no significant effect), while the relative risk for those without thyroid disease is

2.65

(strong significant effect).

Stratification is the primary method used to differentiate between confounding and effect modification.

Compare the stratum-specific risks to each other and to the crude risk.

The strata risks (

1.08

and

2.65

) are significantly different from each other. The crude risk (

1.75

) falls between them.

If strata risks were similar to each other but different from the crude, it would be confounding. Because they differ widely from each other, it is effect modification.

Determine the correct epidemiological response to effect modification.

Effect modification is a natural phenomenon, not an error. It should be highlighted by reporting the stratum-specific results separately.

Adjusting or pooling the data would hide the fact that the drug only works for one specific subset of patients.

Key Concept

Differentiating Effect Modification from Confounding based on stratified analysis

Question

Answer

Step-by-Step Solution

Key Concept

Common Mistakes

Topics

More questions from this topic