A 42-year-old woman from Delhi presents to the outpatient clinic with a 6-month history of progressive dyspnea and cough. She works as a textile factory supervisor and has 15 years of occupational exposure to cotton dust. Chest X-ray shows bilateral lower lobe fibrosis. A researcher wants to investigate the association between cotton dust exposure and the development of byssinosis in textile workers. The researcher identifies 500 workers currently employed in textile factories (250 with byssinosis, 250 without) and retrospectively collects data on their past exposure to cotton dust over the last 20 years. Which study design is most appropriate for this investigation?

Explanation

## Study Design Selection in Occupational Epidemiology ### Why Case-Control is Optimal Here **Key Point:** The researcher has already identified cases (workers with byssinosis) and controls (workers without byssinosis) and is collecting *retrospective* exposure data. This is the defining structure of a case-control study. **High-Yield:** Case-control studies are the design of choice when: - Disease is already established (prevalent or incident cases exist) - Exposure data must be collected retrospectively - The disease is relatively rare or has a long latency period - Resources are limited (smaller sample size than cohort) ### Comparison of Study Designs | Feature | Case-Control | Prospective Cohort | Cross-Sectional | RCT | |---------|--------------|-------------------|-----------------|-----| | **Direction** | Backward (disease → exposure) | Forward (exposure → disease) | Snapshot | Forward | | **Timing** | Retrospective exposure data | Prospective follow-up | Both measured now | Randomized intervention | | **Best for** | Rare diseases, long latency | Common diseases, short latency | Prevalence estimation | Causality proof | | **Odds Ratio** | Directly calculated | Risk Ratio calculated | Prevalence Ratio | Risk Ratio | | **Cost** | Low | High | Low | Very high | ### Why Other Options Are Suboptimal 1. **Prospective Cohort:** Would require following 500 unexposed workers for 15–20 years to develop byssinosis. Impractical, expensive, and the disease has already manifested in current workers. 2. **RCT:** Unethical to randomize workers to high cotton dust exposure. Occupational hazard research cannot use randomization. 3. **Cross-Sectional:** Measures current exposure and current disease, but cannot establish temporal sequence (which came first?). Weaker for causality inference. **Clinical Pearl:** In occupational and environmental epidemiology, when cases already exist and you need to look back at exposure history, case-control is the gold standard — it is efficient and provides an odds ratio that approximates relative risk when the disease is uncommon. **Mnemonic:** **RECALL** — *Retrospective, Already-identified Cases, Comparison group, All exposure data collected backward in time, Low cost, Long latency* = Case-Control design.