641development of predictive genetic tests. Using the example of serious drug-induced
liver injury, a study has shown how a database of routinely collected EHRs can be
used to overcome these barriers by facilitating rapid recruitment to genome-wide
association studies and supporting effi cient randomized controlled trials of predic-
tive genetic test effectiveness (Wing et al. 2014 ).
Use of EHRs for Genetic Research
EHRs are a potential source of longitudinal clinical data for research. The Electronic
Medical Records and Genomics Network (eMERGE), including its partners at
Northwestern, Mayo Clinic, Vanderbilt University, Marshfi eld Clinic Research
Foundation, and Group Health, has investigated whether data captured through rou-
tine clinical care using EHRs can be used to identify disease phenotypes with suf-
fi cient positive and negative predictive values for use in genome-wide association
studies. Using data from 5 different sets of EHRs – type 2 diabetes, dementia,
peripheral arterial disease, cataracts, and cardiac arrhythmia – the investigators
have identifi ed fi ve disease phenotypes with positive predictive values ranging
from 73 % to 98 % and negative predictive values ranging from 98 % to 100 %
(Kho et al. 2011 ). Most EHRs captured key information (diagnoses, medications,
laboratory tests) used to defi ne phenotypes in a structured format. The authors
showed that natural language processing is an important tool for improving case
identifi cation rates. Efforts and incentives to increase the implementation of
interoperable EHRs will markedly improve the availability of clinical data for
genomics research. Researchers studying the underlying genetic causes for human
diseases can dramatically cut their costs and save time by mining data about real
patients found in EHRs, instead of recruiting and sorting participants as they look
for common genetic variants. As the cost of genome sequencing is dropping, it
should eventually be possible to include patients’ genomes in their medical records,
providing a valuable source of information for disease researchers. The larger the
studies, the better they could be at detecting rare effects of genes and providing
more detail about the genetic sequences that lead to diseases. Drawbacks of EHR-
based research are: (1) lack of uniformity as the EHRs used different software; and
(2) EHRs did a poor job of capturing different factors such as race and ethnicity,
smoking status, and family history.
Although the use of EHRs is becoming well established in some health care
systems in the US, such as those of Kaiser Permanente and the Geisinger Health
System, the integration of genetic information is lagging behind as there are 10 dif-
ferent EHR systems. According to the results of a survey of health care profession-
als, only 4 % of the respondents reported that their EHR system provided any
decision support on the basis of the results of genetic tests, and the vast majority
reported that their EHR supplier did not provide the type of support they needed for
the interpretation of genetic information (Scheuner et al. 2009 ).
EHRs could provide a platform for the integration of genetic information into
clinical practice by guiding clinicians about when to order a genetic test, how to
Role of Bioinformatics in Development of Personalized Medicine