One-Year Mortality Risk Estimates from Electronic Medical Record Data

Technology No. 20170266

Random Forest Model Does Not Focus on Specific Disease

An algorithm uses prognostic models based on electronic medical record (EMR) data to predict one-year mortality risk. The random forest model, which predicts the risk of death within a year of the last day of hospitalization, is remarkable as a broadly applicable prognostic model that does not focus on a specific disease. The model is derived from about 100,000 hospital discharges, is validated on about 20,000 discharges and has a C-statistic of about 0.85 to 0.89 (depending on the variables selected) for predicting death within a year from the last day in the hospital. The algorithm uses commonly obtained EMR data (e.g., last set of vital signs, complete blood count, basic and complete metabolic panel, demographic information, ICD codes, etc.), physiological readings (e.g., blood pressure, pulse heart rate, etc.) and basic demographic information (e.g., patient age and length of hospital stay) to estimate the risk of one-year mortality to improve end-of-life planning and risk adjustment for research.

Uses Electronic Health Record Data

Currently, there is no broadly applicable, well-validated model using commonly obtained labs and vitals from last-available hospital data to predict one-year mortality in a heterogeneous set of patients. This new algorithm uses only structured data available in the EMR but also accommodates missing data; therefore, a clinician is not required to enter any data if the model is deployed in an EMR. In addition, a web-based API design is available that allows developers to query the model from anywhere in the world and use it in their applications.


  • Based on electronic medical record (EMR) data
  • Predicts one-year mortality risk within a year of the last day of hospitalization
  • Broadly applicable: does not focus on a specific disease
  • Random forest model
  • Web-based API design available
  • Clinician data input not if deployed in an EMR


  • Predicting one-year mortality risk within a year of the last day of hospitalization
  • Clinical decision support tool (i.e., for end-of-life care)
  • Improves end-of-life planning and risk adjustment for research
  • Web-based API
  • Automated risk estimation
  • Flagging tool

Phase of Development - Prototype

Nishant Sahni, MD, MS

Development and Validation of Machine Learning Models for Prediction of 1-Year Mortality Utilizing Electronic Medical Record Data Available at the End of Hospitalization in Multicondition Patients: a Proof-of-Concept Study
Journal of General Internal Medicine, Jan 30, 2018; pp 1-8

Interested in Licensing?
The University relies on industry partners to scale up technologies to large enough production capacity for commercial purposes. The license is available for this technology and would be for the sale, manufacture or use of products claimed by the issued patents. Please contact us to share your business needs and technical interest in this technology and if you are interested in licensing the technology for further research and development.
  • swap_vertical_circlecloud_downloadSupporting documents (0)
Questions about this technology?