A 61 year old woman with chronic obstructive pulmonary disease enrolls in a cohort study of coffee intake and lung cancer. Investigators adjust for smoking status as current, former, or never smoker. After adjustment, heavy coffee intake still appears associated with lung cancer, but the authors note that pack years were not measured. Which of the following best explains the concern that smoking may still distort the association?
- A)Differential misclassification
- B)Interaction
- C)Type I error
- D)Selection bias
- E)Residual confoundingGABARITO
Explicação
Residual confounding persists when an important confounder is measured imprecisely or categorized too crudely for adequate adjustment. Smoking is strongly associated with both coffee use and lung cancer, and adjustment only by current, former, or never status ... Ver explicação completa e trilha adaptativa →