What question did this study set out to answer?

The aim is to improve the assessment of risk models' accuracy when some predictors are missing by using auxiliary information.

Incorporating Auxiliary Information into Assessment of Accuracy and Discrimination of Risk Models When Some Predictors are Missing

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial evaluates risk model accuracy by adjusting weights using auxiliary information, suggesting enhanced validation strategies.

The aim is to improve the assessment of risk models' accuracy when some predictors are missing by using auxiliary information.
Adjusted known sampling weights by incorporating auxiliary information for all cohort members.
Assessed measures of classification accuracy, including true and false positive rates and AUC.
Compared efficiency of weight adjustment with multiple imputation in simulations.
Weight adjustment using auxiliary information improved efficiency of risk model estimates compared to standard approaches.
Multiple imputation often yielded biased performance estimates under mis-specified models.
Illustrated techniques with an absolute risk model for second primary thyroid cancer in an independent cohort.

Pfeiffer et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark

Randomized trial evaluates risk model accuracy by adjusting weights using auxiliary information, suggesting enhanced validation strategies.