This study developed and evaluated methods to adjust misclassification errors in electronic health record (EHR)-derived covariates using group-wise and individualized weights based on observed sensitivity and specificity to reduce bias in predictive modeling. ... ...

请注册登录后继续浏览