AUTHOR=Wang Liwei , Olson Janet E. , Bielinski Suzette J. , St. Sauver Jennifer L. , Fu Sunyang , He Huan , Cicek Mine S. , Hathcock Matthew A. , Cerhan James R. , Liu Hongfang 

TITLE=Impact of Diverse Data Sources on Computational Phenotyping

JOURNAL=Frontiers in Genetics

VOLUME=Volume 11 - 2020

YEAR=2020

URL=https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2020.00556

DOI=10.3389/fgene.2020.00556

ISSN=1664-8021

ABSTRACT=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 recall of only 69% and 72% in Mayo data, respectively, 83% and 97% in REP. RA and T2DM controls also contain biases. Recall of RA controls were slightly affected with performance of 99.8% for Mayo and REP respectively, while precision of RA controls was more seriously affected with 67% and 97% for Mayo and REP respectively. Precision of T2DM controls were 91% and 85% for Mayo and REP respectively. We further elaborated underlying reasons impacting the performance.