How Artificial Intelligence is affecting electronic medical records systems

In 2009, the American Recovery and Reinvestment Act (ARRA) spurred significant healthcare and life sciences research, as part of the government’s response to the economic recession. ARRA gave hospitals and medical centers the opportunity (and incentive) to invest in their information technology infrastructure—namely through the Health Information Technology for Economic and Clinical Health (HITECH) Act1. As a result, adoption of electronic medical records systems climbed from ~11% in 2006 to ~96% in 20172.

The impact that the digitization of health records and other clinical data has had on patient management can’t be understated3. Developed initially as billing programs, nearly every aspect of the physician’s workflow has been impacted and changed due to electronic health records systems (EHRs). Long gone are the color-coded paper records found in doctors’ offices only a decade ago, replaced by in-room computers and mobile workstations.

In MoneyBall Medicine: Thriving in the New Data-Driven Healthcare Market, I described how EHRs play a significant role in healthcare. Prescription and lab orders are sent electronically to pharmacies and labs, while built-in algorithms call attention to potential drug interactions—before the patient even takes the medication. Clinical decision support systems, such as UpToDate, put vast amounts of information at doctors’ fingertips and guide patient management to align with clinical guidelines. At several academic medical centers, de-identified EHRs are used to accelerate pharmaceutical and diagnostic development and for population health research. Companies like Nashville Biosciences are leveraging these unique datasets and forging partnerships between industry, academia, and healthcare.

But the growing use of machine learning and artificial intelligence in healthcare shows special promise when combined with the data in EHRs. These technologies require massive amounts of data before they can accurately be used for tasks such as predicting which patients are at risk for hospital readmissions4 or who is developing sepsis5. The rapidly changing needs of EHR users present numerous opportunities for existing EHR developers, such as Cerner and Epic, as well as for startups and current information technology companies looking to create programs that will help doctors with patient management and clinical workflow as well as researchers looking to use AI and predictive analytics.

In Episode 5 of the MoneyBall Medicine Podcast my guest, John Glaser, Senior Vice-President of Population Health at Cerner, and I discuss the evolution of health information technology and the growing role of artificial intelligence and machine learning. We talk about the current paradigm shift from fee-for-service to value-based care and how mergers, consolidations, and partnerships are bringing new types of companies into healthcare. Join us for the next episode of the MoneyBall Medicine Podcast!

  1. Vestal, C. 2014. Some states lag in using electronic health records. USA Today. https://www.usatoday.com/story/news/nation/2014/03/19/stateline-electronic-health-records/6600377/
  2. The Office of the National Coordinator for Health Information Technology. 2018. Non-federal Acute Care Hospital Electronic Health Record Adoption. https://dashboard.healthit.gov/quickstats/pages/FIG-Hospital-EHR-Adoption.php
  3. Health and Human Services. 2018. HITECH Act Enforcement Interim Final Rule. https://www.hhs.gov/hipaa/for-professionals/special-topics/hitech-act-enforcement-interim-final-rule/index.html
  4. Pennic, F. 2017.Partners Connected Health to Develop AI Tool to Predict Risk of Hospital Readmissions. HIT Consultant. https://hitconsultant.net/2017/12/12/partners-connected-health-ai-hospital-readmissions/
  5. Kent, J. 2018.Machine Learning, EHR Big Data Analytics Predict Sepsis. Health IT Analytics. https://healthitanalytics.com/news/machine-learning-ehr-big-data-analytics-predict-sepsis