Prevalence of Heart Failure Signs and Symptoms in a Large Primary Care Population Identified Through the Use of Text and Data Mining of the Electronic Health Record

摘要: Abstract Background The electronic health record (EHR) contains a tremendous amount of data that if appropriately detected can lead to earlier identification disease states such as heart failure (HF). Using novel text and analytic tool we explored the longitudinal EHR over 50,000 primary care patients identify documentation signs symptoms HF in years preceding its diagnosis. Methods Results Retrospective analysis consisted 4,644 incident cases 45,981 group-matched control subjects. Documentation Framingham within encounter notes were carried out with use previously validated natural language processing procedure. A total 892,805 affirmed criteria documented an average observation period 3.4 years. Among eventual cases, 85% had ≥1 criterion 1 year before their diagnosis, did 55% Substantial variability prevalence individual found both case Conclusions are frequently population identified through automated mining EHRs. Their frequent demonstrates rich available EHRs will allow for future work on help develop predictive models HF.