"In a retrospective controlled cohort study, if ~25% of participants have missing data on a crucial variable, is it acceptable to run an adjusted model by excluding these participants? Should the unadjusted model only include participants with no missing values?"
Hussain Nizam
Preeti Maharjan
In a retrospective controlled cohort study, handling missing data is crucial to ensure the validity and reliability of the results. Here are some considerations:
- Adjusted Model: Running an adjusted model by excluding participants with missing data (complete-case analysis) is common. Still, it can lead to biased results if the missing data are not completely random and reduces the sample size, decreasing the study's statistical power.
- Unadjusted Model: The unadjusted model should ideally include all participants, but if those with missing values are excluded, it should be clearly stated, as this can also lead to biased results if the missing data are not random.
- Other Methods: More sophisticated methods to handle missing data, such as multiple imputation or likelihood-based methods, can provide more accurate and unbiased estimates by accounting for the uncertainty associated with missing data.
While excluding participants with missing data is a common practice, it is not always the best approach. Considering alternative methods and clearly reporting the handling of missing data can improve the validity of study results.
1 votes
0 thanks
- Badges
- Science topic
- Similar topics
- Correction
More Preeti Maharjan's questions See All
Similar questions and discussions