A public health researcher in Delhi is investigating the association between occupational dust exposure and chronic obstructive pulmonary disease (COPD) among construction workers. The researcher has identified 500 workers with COPD and 500 workers without COPD, and plans to retrospectively collect data on past dust exposure. What is the most appropriate next step to strengthen the validity of this study design?

Question

Accepted Answer

B. Match cases and controls on age, smoking status, and duration of employment to reduce confounding. ## Study Design Recognition

**Key Point:** This is a **case-control study** — cases (COPD+) and controls (COPD−) are identified first, and exposure history is collected retrospectively. The researcher must address confounding and selection bias.

## Why Matching is the Correct Next Step

**High-Yield:** Matching cases and controls on known confounders (age, smoking status, employment duration) is the **gold standard** for controlling confounding in case-control studies at the design stage. This reduces bias and increases the precision of the odds ratio estimate.

**Clinical Pearl:** In occupational epidemiology, smoking is a major confounder in COPD studies. Matching ensures that the observed association between dust exposure and COPD is not spuriously inflated by differences in smoking prevalence between cases and controls.

## Why Other Options Are Suboptimal

| Option | Problem |
|--------|----------|
| **Recruit only currently employed** | Introduces **survivor bias** — workers who left due to COPD-related disability will be excluded, underestimating the true exposure–disease association. |
| **Exclude all smokers** | Overrestricts the study population, reducing generalizability and statistical power. Matching is preferable to exclusion. |
| **Increase sample size alone** | Does not address confounding; larger sample size with unmatched controls will still yield a biased estimate. |

## Mnemonic: MATCH in Case-Control

**M**atching reduces **C**onfounding  
**A**djustment in analysis (alternative)  
**T**emperament of design (stratification)  
**C**ontrol selection (careful)  
**H**igh precision, lower bias

[cite:Park 26e Ch 10]

Answer

A. Exclude all smokers from both cases and controls to eliminate smoking as a confounder

Answer

C. Increase the sample size to 2000 cases and 2000 controls to improve statistical power

Answer

D. Recruit only workers currently employed to ensure accurate exposure data collection

Explanation

Practice similar questions