Examkin

Question

Difficulty: MediumBias and Confounding in Research

A research group is investigating the association between a newly approved anti-obesity medication and the incidence of acute cholecystitis. In an initial prospective observational cohort study, the crude relative risk of acute cholecystitis in patients taking the new medication compared to those on standard lifestyle interventions is 2.2 (95% CI: 1.6-2.9).

During data review, the investigators observe that the new medication was disproportionately prescribed to patients with a baseline body mass index (BMI) > 40 kg/m². They subsequently stratify the data by baseline BMI. In the stratum of patients with a BMI > 40 kg/m², the relative risk is 1.1 (95% CI: 0.8-1.5). In the stratum of patients with a BMI ≤ 40 kg/m², the relative risk is also 1.1 (95% CI: 0.9-1.4).

The research group is now planning a subsequent observational cohort study to further investigate this relationship. They want to ensure that baseline BMI does not distort the association, and they specifically wish to address this issue during the design phase of the new study.

Which of the following methods is most appropriate to achieve this goal?

  1. Matching exposed and unexposed subjects by baseline body mass indexAnswer
  2. B
    Adjusting for body mass index using multivariable logistic regression
  3. C
    Performing a stratified analysis by body mass index categories
  4. D
    Selecting study participants randomly from a national health registry
  5. E
    Increasing the overall study sample size to enhance statistical power

Answer

Matching exposed and unexposed subjects by baseline body mass index
The scenario describes confounding, evidenced by a crude relative risk that differs from the strata-specific relative risks. The prompt specifically asks for a method to control for confounding during the design phase of an observational study. Matching is a design-stage method where investigators intentionally select participants such that the potential confounder is equally distributed between the exposed and unexposed groups. Other design-phase methods include restriction (limiting study entry criteria) and randomization (applicable only to experimental trials, not observational studies).

Step-by-Step Solution

1
Analyze the initial study data.
The crude relative risk (2.2) differs significantly from the stratified relative risks (both 1.1).
This confirms that baseline BMI is a confounder, specifically demonstrating confounding by indication.
2
Identify the researchers' goal for the new study.
The researchers want to control for this confounding variable specifically during the design phase of an observational study.
Different methods exist for controlling confounding, but they are strictly divided into design-phase and analysis-phase techniques.
3
Evaluate the options based on study phase.
Matching is a design-phase technique. Regression and stratification are analysis-phase techniques.
Selecting a design-phase method directly answers the question prompt.

Key Concept

Methods to control for confounding are categorized into design-phase (randomization, matching, restriction) and analysis-phase (stratification, multivariable regression).
Estimated Time:1m 15s
Rate this question

Topics

Bias and Confounding in Research

More questions from this topic