English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Background: Existing guidelines for handling missing data are generally notconsistent with the goals of prediction modelling, where missing data can occurat any stage of the model pipeline. Multiple imputation (MI), often heralded asthe gold standard approach, can be challenging to apply in the clinic. Clearly,the outcome cannot be used to impute data at prediction time. Regressionimputation (RI) may offer a pragmatic alternative in the prediction context,that is simpler to apply in the clinic. Moreover, the use of missing indicatorscan handle informative missingness, but it is currently unknown how well theyperform within CPMs. Methods: We performed a simulation study where data weregenerated under various missing data mechanisms to compare the predictiveperformance of CPMs developed using both imputation methods. We considerdeployment scenarios where missing data is permitted/prohibited, and developmodels that use/omit the outcome during imputation and include/omit missingindicators. Results: When complete data must be available at deployment, ourfindings were in line with widely used recommendations; that the outcome shouldbe used to impute development data under MI, yet omitted under RI. Whenimputation is applied at deployment, omitting the outcome from the imputationat development was preferred. Missing indicators improved model performance insome specific cases, but can be harmful when missingness is dependent on theoutcome. Conclusion: We provide evidence that commonly taught principles ofhandling missing data via MI may not apply to CPMs, particularly when data canbe missing at deployment. In such settings, RI and missing indicator methodscan (marginally) outperform MI. As shown, the performance of the missing datahandling method must be evaluated on a study-by-study basis, and should bebased on whether missing data are allowed at deployment.

Imputation and Missing Indicators for handling missing data in the development and implementation of clinical prediction models: a simulation study