JMIR Publications

ABSTRACT

Background:

Sepsis is a life-threatening syndrome that can rapidly lead to organ damage and death. Existing risk scores predict outcomes for patients who have already become acutely ill.

Objective:

Our objective is to develop a model to identify patients at risk of getting sepsis within two years, to support reduction of sepsis morbidity and mortality.

Methods:

Machine learning was applied to 2,683,049 electronic health records (EHRs) with over 64 million encounters across five states to develop models for predicting a patient’s risk of getting sepsis within two years. Features were selected to be easily obtainable from a patient’s chart in real-time during ambulatory encounters.

Results:

Models showed consistent prediction scores, with the highest AUROC of 0.82 and positive likelihood ratio of 2.9 achieved with gradient boosting on all features combined. Predictive features included age, sex, ethnicity, average ambulatory heart rate, standard deviation of body mass index, and the number of prior medical conditions and procedures. Results identified both known and potential new risk factors for long-term sepsis. Model variations also illustrated trade-offs between incrementally higher accuracy, implementability and interpretability.

Conclusions:

Accurate, implementable models were developed to predict two-year risk of sepsis, using EHR data that is easy to obtain from ambulatory encounters. These results help advance understanding of sepsis and provide a foundation for future trials of risk-informed preventive care.