Compatibility in Missing Data Handling Across the Prediction Model Pipeline: A Simulation Study

被引：0

作者：

Tsvetanova, Antonia ^{[1
]}

Sperrin, Matthew ^{[1
]}

Jenkins, David ^{[1
]}

Peek, Niels ^{[1
,2
]}

Buchan, Iain ^{[3
]}

Hyland, Stephanie ^{[4
]}

Martin, Glen ^{[1
]}

机构：

[1] Univ Manchester, Ctr Hlth Informat, Fac Biol Med & Hlth, Manchester, England

[2] Univ Manchester, Christabel Pankhurst Inst Hlth Technol Res & Inno, Manchester, England

[3] Univ Liverpool, Inst Populat Hlth, Liverpool, England

[4] Microsoft Res Cambridge, Cambridge, England

来源：

MEDINFO 2023 - THE FUTURE IS ACCESSIBLE | 2024年 / 310卷

关键词：

Statistical models; missing data; imputation; simulation study;

D O I：

10.3233/SHTI231252

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Careful handling of missing data is crucial to ensure that clinical prediction models are developed, validated, and implemented in a robust manner. We determined the bias in estimating predictive performance of different combinations of approaches for handling missing data across validation and implementation. We found four strategies that are compatible across the model pipeline and have provided recommendations for handling missing data between model validation and implementation under different missingness mechanisms.

引用

页码：1476 / 1477

页数：2