[Skip to Content]
[Skip to Content Landing]
Invited Commentary
February 21, 2019

Data Analytics and Machine Learning for Disease Identification in Electronic Health Records

Author Affiliations
  • 1School of Engineering, University of Melbourne, Parkville, Victoria, Australia
  • 2Centre for AgriBiosciences, AgriBio, State Government of Victoria, Bundoora, Victoria, Australia
JAMA Ophthalmol. 2019;137(5):497-498. doi:10.1001/jamaophthalmol.2018.7055

In this issue of JAMA Ophthalmology, Stein et al1 reported that nearly 75% of encounters between patients and eye care practitioners are documented in the form of electronic health record (EHR) data that provide details of test results, historical information, management plans, and billing codes. This emerging large-scale database provides researchers with so-called big data to support the development of algorithms for the detection of disease or the presence or absence of a range of medical conditions.2,3 A topic of research attracting increasing interest is the application of data science and predictive analytics to develop generalized algorithms for the detection of phenotypes of interest in EHR data and other large-scale data sets.