Publications

BibTex format

@article{Denaxas:2020:jamiaopen/ooaa047,
author = {Denaxas, S and Shah, AD and Mateen, BA and Kuan, V and Quint, J and Fitzpatrick, N and Torralbo, A and Fatemifar, G and Hemingway, H},
doi = {jamiaopen/ooaa047},
journal = {Jama Network Open},
pages = {545--556},
title = {A semi-supervised approach for rapidly creating clinical biomarker phenotypes in the UK Biobank using different primary care EHR and clinical terminology systems},
url = {http://dx.doi.org/10.1093/jamiaopen/ooaa047},
volume = {3},
year = {2020}
}

Download

RIS format (EndNote, RefMan)

TY  - JOUR
AB  - ObjectivesThe UK Biobank (UKB) is making primary care electronic health records (EHRs) for 500 000 participants available for COVID-19-related research. Data are extracted from four sources, recorded using five clinical terminologies and stored in different schemas. The aims of our research were to: (a) develop a semi-supervised approach for bootstrapping EHR phenotyping algorithms in UKB EHR, and (b) to evaluate our approach by implementing and evaluating phenotypes for 31 common biomarkers.Materials and MethodsWe describe an algorithmic approach to phenotyping biomarkers in primary care EHR involving (a) bootstrapping definitions using existing phenotypes, (b) excluding generic, rare, or semantically distant terms, (c) forward-mapping terminology terms, (d) expert review, and (e) data extraction. We evaluated the phenotypes by assessing the ability to reproduce known epidemiological associations with all-cause mortality using Cox proportional hazards models.ResultsWe created and evaluated phenotyping algorithms for 31 biomarkers many of which are directly related to COVID-19 complications, for example diabetes, cardiovascular disease, respiratory disease. Our algorithm identified 1651 Read v2 and Clinical Terms Version 3 terms and automatically excluded 1228 terms. Clinical review excluded 103 terms and included 44 terms, resulting in 364 terms for data extraction (sensitivity 0.89, specificity 0.92). We extracted 38 190 682 events and identified 220 978 participants with at least one biomarker measured.Discussion and conclusionBootstrapping phenotyping algorithms from similar EHR can potentially address pre-existing methodological concerns that undermine the outputs of biomarker discovery pipelines and provide research-quality phenotyping algorithms.
AU  - Denaxas,S
AU  - Shah,AD
AU  - Mateen,BA
AU  - Kuan,V
AU  - Quint,J
AU  - Fitzpatrick,N
AU  - Torralbo,A
AU  - Fatemifar,G
AU  - Hemingway,H
DO  - jamiaopen/ooaa047
EP  - 556
PY  - 2020///
SN  - 2574-3805
SP  - 545
TI  - A semi-supervised approach for rapidly creating clinical biomarker phenotypes in the UK Biobank using different primary care EHR and clinical terminology systems
T2  - Jama Network Open
UR  - http://dx.doi.org/10.1093/jamiaopen/ooaa047
UR  - http://hdl.handle.net/10044/1/83052
VL  - 3
ER  -

Download

ProfessorJenniferQuint

Contact

Location

Summary

Citation

BibTex format

RIS format (EndNote, RefMan)