A research team conducts a retrospective cohort study to evaluate the risk of acute kidney… — USMLE Step 3

A research team conducts a retrospective cohort study to evaluate the risk of acute kidney injury associated with a novel immunosuppressant in lung transplant recipients. The primary analysis yields a crude relative risk ( $RR$ ) of $2.8$ ( $95\% \text{ CI}: 1.9 - 4.1$ ) for acute kidney injury in patients receiving the novel agent compared to those on standard therapy. Concerned about baseline patient differences, the investigators perform a stratified analysis based on the presence or absence of pre-existing systemic hypertension. The stratum-specific relative risk is $1.1$ ( $95\% \text{ CI}: 0.7 - 1.6$ ) in patients with pre-existing hypertension and $1.2$ ( $95\% \text{ CI}: 0.8 - 1.7$ ) in patients without pre-existing hypertension.

Which of the following is the most accurate conclusion regarding the role of pre-existing hypertension in this study?

It is a confounding variable that artificially inflated the apparent association in the initial analysis.Answer
B
It is an effect modifier that synergistically interacts with the novel immunosuppressant to cause acute kidney injury.
C
It indicates that a chi-square test must be replaced with an analysis of variance (ANOVA) to compare the incidence across strata.
D
It demonstrates that the positive predictive value of the novel immunosuppressant for acute kidney injury varies with hypertension prevalence.
E
It suggests that the initial findings were likely due to isolated, non-diagnostic fluctuations in serum creatinine rather than true injury.

Answer

It is a confounding variable that artificially inflated the apparent association in the initial analysis.

The correct answer accurately identifies the statistical phenomenon of confounding. A confounding variable is an extraneous factor that correlates with both the dependent and independent variables. In a stratified analysis, confounding is present when the stratum-specific measures of association (e.g., Relative Risk, Odds Ratio) are similar to each other but differ substantially from the crude, unadjusted measure. In this scenario, the crude

RR

2.8

dropped to non-significant values of

1.1

and

1.2

after stratifying by hypertension. This indicates that hypertension—not the novel drug—was driving the increased risk of acute kidney injury in the initial analysis.

Step-by-Step Solution

Analyze the crude relative risk (

RR

) provided in the study.

The crude

RR

2.8

, suggesting a strong, statistically significant association between the novel immunosuppressant and acute kidney injury.

Establishing the baseline finding before stratification is necessary to compare against the adjusted findings.

Evaluate the stratum-specific relative risks after adjusting for pre-existing hypertension.

The stratum-specific

RR

values are

1.1

and

1.2

. Neither is statistically significant (both

95\% \text{ CI}

s include

1.0

), and they are similar to each other.

Comparing stratum-specific values determines how the third variable interacts with the exposure and outcome.

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

The stratum-specific relative risks (

1.1

and

1.2

) are similar to each other but notably different from the crude relative risk (

2.8

This pattern is the hallmark of confounding. If the stratum-specific values were significantly different from each other, it would indicate effect modification.

Key Concept

In epidemiological studies, confounding occurs when the crude measure of association is significantly different from the stratum-specific measures of association, which are similar to each other. Effect modification occurs when the stratum-specific measures are significantly different from each other.

Question

Answer

Step-by-Step Solution

Key Concept

Common Mistakes

Topics

More questions from this topic