A epidemiologic study examining 5,000 patients (mean age 62, 55% male) finds coffee consumption associated with lung cancer (RR 2.1, p<0.05). Chest CT confirms adenocarcinoma in 340 subjects. When data are stratified by smoking status, the association disappears in both smokers and nonsmokers. Notably, no dose-response relationship exists between coffee quantity and cancer risk. Which of the following best explains the original association?
- A)ConfoundingGABARITO
- B)Effect modification
- C)Selection bias
- D)Type I error
- E)Lead-time bias
Explicação
Confounding occurs when a third variable is associated with both the exposure and the outcome but is not on the causal pathway. Smoking is associated with coffee drinking and independently causes lung cancer, so failing to account for smoking creates a spuriou... Ver explicação completa e trilha adaptativa →