A team of public health researchers is designing an observational case-control study to evaluate the relationship between artificial sweetener use and the development of bladder cancer. They identify cigarette smoking as a major potential confounder in this relationship. Which of the following methods can be appropriately utilized during the design phase of this study to minimize the confounding effect of smoking? (Select all that apply)
- AStratifying the collected data by smoking status before calculating the final odds ratio
- Matching cases and controls based on their smoking historyAnswer
- CApplying multivariable logistic regression to mathematically adjust for smoking
- DIncreasing the overall target sample size of the study population
- Restricting study enrollment exclusively to individuals who have never smokedAnswer
Answer
The correct methods to utilize during the design phase are matching cases and controls based on their smoking history, and restricting study enrollment exclusively to non-smokers.
To prevent confounding before data is even collected, researchers must use design-phase strategies. Matching ensures that for every case enrolled, a control with the exact same confounding trait (e.g., smoking status) is also enrolled, forcing equal distribution. Restriction simply eliminates the confounder entirely by only enrolling participants from one category (e.g., only non-smokers). Both methods are proactively implemented during study design.
Step-by-Step Solution
Key Concept
Strategies to control confounding are divided into design-stage methods (randomization, matching, restriction) and analysis-stage methods (stratification, multivariable regression).