Examkin

Question

Difficulty: HardBias and Confounding in Research

Researchers conduct a retrospective cohort study to evaluate the risk of severe hypertensive episodes associated with a newly approved monoclonal antibody targeting the calcitonin gene-related peptide (CGRP) pathway. The cohort includes 10,000 adult patients with chronic migraine. The primary outcome is the occurrence of a severe hypertensive episode within 6 months of initiating therapy. After initial data collection, the researchers stratify the results based on the patients' concurrent use of triptans as abortive therapy. The findings are summarized in the table below:

Study GroupRelative Risk (95% CI)
Crude (Overall Cohort)1.6 (1.2 - 2.1)
Stratum 1: Concurrent Triptan Use3.2 (2.1 - 4.8)
Stratum 2: No Concurrent Triptan Use1.05 (0.8 - 1.3)

Based on these findings, which of the following is the most accurate description of the role of concurrent triptan use in this analysis?

  1. It acts as an effect modifier, and the stratum-specific relative risks should be reported instead of the crude estimateAnswer
  2. B
    It is a confounding variable, and an adjusted relative risk should be calculated using multivariable regression
  3. C
    It is a confounding variable, which should have been controlled for during the design phase through matching
  4. D
    It introduces selection bias, and the study must be restricted solely to patients not using triptans
  5. E
    It causes observer bias, as investigators are more likely to monitor patients taking multiple medications for hypertension

Answer

It acts as an effect modifier, and the stratum-specific relative risks should be reported instead of the crude estimate.
The correct answer accurately identifies the statistical phenomenon as effect modification (also known as interaction). Effect modification occurs when the magnitude of the effect of an exposure on an outcome differs depending on the level of a third variable. This is assessed analytically via stratification. If the relative risks (or odds ratios) in the different strata are significantly different from one another (e.g., 3.2 vs 1.05), effect modification is present. Because the exposure has a different effect depending on the presence of this third variable, it is inappropriate to calculate an 'adjusted' overall risk or report the crude risk; instead, researchers should report the stratum-specific risks to accurately reflect the data.

Step-by-Step Solution

1
Examine the crude measure of association.
The overall crude Relative Risk (RR) is 1.6, suggesting a significant association between the CGRP inhibitor and severe hypertensive episodes.
Establishing the baseline overall risk is necessary before evaluating the impact of stratification.
2
Analyze the stratum-specific measures of association.
The RR is 3.2 for triptan users and 1.05 for non-triptan users.
Stratification separates the cohort into subgroups to test for confounding and effect modification.
3
Compare the stratum-specific risks to each other and to the crude risk.
The stratum-specific RRs (3.2 and 1.05) are significantly different from each other.
When stratum-specific estimates differ significantly from one another, it indicates that the effect of the primary exposure depends on the presence of the third variable. This is the definition of effect modification.
4
Determine the appropriate reporting method.
Because effect modification is present, reporting the crude or adjusted RR would be misleading. The stratum-specific RRs must be reported.
An adjusted or crude estimate obscures the biological reality that the medication only increases risk in a specific subgroup.

Key Concept

Distinguishing effect modification from confounding based on stratified analysis
Rate this question

Topics

Bias and Confounding in Research

More questions from this topic