A platform for phenotyping disease progression and associated longitudinal risk factors in large-scale EHRs, with application to incident diabetes complications in the UK Biobank. Issue 1 (9th February 2023)

• Record Type: Journal Article
• Authors: Kim, Do Hyun; Jensen, Aubrey; Jones, Kelly; Raghavan, Sridharan; Phillips, Lawrence S; Hung, Adriana; Sun, Yan V; Li, Gang; Reaven, Peter; Zhou, Hua; Zhou, Jin J
• Is Part Of: JAMIA open. Volume 6:Issue 1(2023)
• Publication Date: 2023-02-09
• Subjects: phenotyping -- diabetes -- diabetes complications -- disease progression -- electronic health records -- time-to-event; Medical informatics -- Periodicals; 610.285
• DOI: 10.1093/jamiaopen/ooad006
• Languages: English
• ISSNs: 2574-2531

Abstract

Abstract: Objective: Modern healthcare data reflect massive multi-level and multi-scale information collected over many years. The majority of the existing phenotyping algorithms use case–control definitions of disease. This paper aims to study the time to disease onset and progression and identify the time-varying risk factors that drive them. Materials and Methods: We developed an algorithmic approach to phenotyping the incidence of diseases by consolidating data sources from the UK Biobank (UKB), including primary care electronic health records (EHRs). We focused on defining events, event dates, and their censoring time, including relevant terms and existing phenotypes, excluding generic, rare, or semantically distant terms, forward-mapping terminology terms, and expert review. We applied our approach to phenotyping diabetes complications, including a composite cardiovascular disease (CVD) outcome, diabetic kidney disease (DKD), and diabetic retinopathy (DR), in the UKB study. Results: We identified 49 049 participants with diabetes. Among them, 1023 had type 1 diabetes (T1D), and 40 193 had type 2 diabetes (T2D). A total of 23 833 diabetes subjects had linked primary care records. There were 3237, 3113, and 4922 patients with CVD, DKD, and DR events, respectively. The risk prediction performance for each outcome was assessed, and our results are consistent with the prediction area under the ROC (receiver operating characteristic) curve (AUC) of standard risk prediction models using cohort studies. Discussion and Conclusion: Our publicly available pipeline and platform enable streamlined curation of incidence events, identification of time-varying risk factors underlying disease progression, and the definition of a relevant cohort for time-to-event analyses. These important steps need to be considered simultaneously to study disease progression. Lay Summary: A modern biobank-scale health care data linked with electronic health care records such as UK Biobank (UKB) provide researchers with a tremendous opportunity to study disease progression and associated risk factors. To study disease progression, researchers often use time-to-event statistical analysis. However, time-to-event data are not readily available from primary sources of healthcare data and require concerted efforts to create them. In this paper, we introduce modularized procedures for systematically phenotyping time-to-event outcomes at large scale. We consistently and accurately curate biomarker trajectory data and define relevant disease outcomes and event times. Adapting our procedures to datasets from the UKB study and using diabetes-related complications as examples, we curate trajectory data of diabetes risk factors and phenotype time to the onset of diabetes vascular complications (ie, cardiovascular complications, diabetic kidney disease, and diabetic retinopathy). This allows us to assess the effects of several known risk factors for the onset of diabetes complications. A risk prediction analysis shows prediction accuracy consistent with the ones obtained from standard risk prediction models using cohort studies.