event-icon
Description

Abstract: EHRs have been proven a valuable resource for conducting large-scale clinical researches. Nonetheless, missing data is very common in EHR, leading to insufficient information. As a result, investigators often need to filter patients by study-specific criteria on data completeness, which may introduce potential biases and subsequently affect the downstream analyses. This study aims to assess the effect of biases introduced by different data completeness filters on predicting diabetic kidney disease (DKD) risk among diabetic patients.

Learning Objective 1: This study aims to investigate the effect of different data completeness filters on the study cohort and subsequent prediction of diabetic kidney disease (DKD) risk among patients with type 2 Diabetes Mellitus (DM).

Authors:

Xing Song (Presenter)
University of Kansas Medical Center

Russ Waitman, University of Kansas Medical Center
Yong Hu, Jinan University
Alan Yu, University of Kansas Medical Center
David Robins, University of Kansas Medical Center
Mei Liu, University of Kansas Medical Center

Presentation Materials:

Keywords