TY - JOUR
T1 - Impact of Diverse Data Sources on Computational Phenotyping
AU - Wang, Liwei
AU - Olson, Janet E.
AU - Bielinski, Suzette J.
AU - St. Sauver, Jennifer L.
AU - Fu, Sunyang
AU - He, Huan
AU - Cicek, Mine S.
AU - Hathcock, Matthew A.
AU - Cerhan, James R.
AU - Liu, Hongfang
N1 - Funding Information:
We acknowledge the Mayo Clinic Center for Individualized Medicine. We acknowledge Donna M. Ihrke for manual chart review. Funding. This work was supported by the Mayo internal funding. This work was also supported by the National Center for Advancing Translational Sciences (U01TR02062) and National Library of Medicine of the National Institutes of Health under Award Number R01LM011829.
Publisher Copyright:
© Copyright © 2020 Wang, Olson, Bielinski, St. Sauver, Fu, He, Cicek, Hathcock, Cerhan and Liu.
PY - 2020/6/3
Y1 - 2020/6/3
N2 - Electronic health records (EHRs) are widely adopted with a great potential to serve as a rich, integrated source of phenotype information. Computational phenotyping, which extracts phenotypes from EHR data automatically, can accelerate the adoption and utilization of phenotype-driven efforts to advance scientific discovery and improve healthcare delivery. A list of computational phenotyping algorithms has been published but data fragmentation, i.e., incomplete data within one single data source, has been raised as an inherent limitation of computational phenotyping. In this study, we investigated the impact of diverse data sources on two published computational phenotyping algorithms, rheumatoid arthritis (RA) and type 2 diabetes mellitus (T2DM), using Mayo EHRs and Rochester Epidemiology Project (REP) which links medical records from multiple health care systems. Results showed that both RA (less prevalent) and T2DM (more prevalent) case selections were markedly impacted by data fragmentation, with positive predictive value (PPV) of 91.4 and 92.4%, false-negative rate (FNR) of 26.6 and 14% in Mayo data, respectively, PPV of 97.2 and 98.3%, FNR of 5.2 and 3.3% in REP. T2DM controls also contain biases, with PPV of 91.2% and FNR of 1.2% for Mayo. We further elaborated underlying reasons impacting the performance.
AB - Electronic health records (EHRs) are widely adopted with a great potential to serve as a rich, integrated source of phenotype information. Computational phenotyping, which extracts phenotypes from EHR data automatically, can accelerate the adoption and utilization of phenotype-driven efforts to advance scientific discovery and improve healthcare delivery. A list of computational phenotyping algorithms has been published but data fragmentation, i.e., incomplete data within one single data source, has been raised as an inherent limitation of computational phenotyping. In this study, we investigated the impact of diverse data sources on two published computational phenotyping algorithms, rheumatoid arthritis (RA) and type 2 diabetes mellitus (T2DM), using Mayo EHRs and Rochester Epidemiology Project (REP) which links medical records from multiple health care systems. Results showed that both RA (less prevalent) and T2DM (more prevalent) case selections were markedly impacted by data fragmentation, with positive predictive value (PPV) of 91.4 and 92.4%, false-negative rate (FNR) of 26.6 and 14% in Mayo data, respectively, PPV of 97.2 and 98.3%, FNR of 5.2 and 3.3% in REP. T2DM controls also contain biases, with PPV of 91.2% and FNR of 1.2% for Mayo. We further elaborated underlying reasons impacting the performance.
KW - computational phenotyping
KW - diverse data sources
KW - phenotyping algorithms
KW - rheumatoid arthritis
KW - type 2 diabetes mellitus
UR - http://www.scopus.com/inward/record.url?scp=85086785834&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85086785834&partnerID=8YFLogxK
U2 - 10.3389/fgene.2020.00556
DO - 10.3389/fgene.2020.00556
M3 - Article
AN - SCOPUS:85086785834
SN - 1664-8021
VL - 11
JO - Frontiers in Genetics
JF - Frontiers in Genetics
M1 - 556
ER -