Online Brief ReportsUtilization of the Signature Method to Identify the Early Onset of Sepsis From Multivariate Physiological Time Series in Critical Care MonitoringMorrill, James H. MMath1,2; Kormilitzin, Andrey PhD1,3; Nevado-Holgado, Alejo J. PhD3; Swaminathan, Sumanth PhD2,4; Howison, Samuel D. DPhil1; Lyons, Terry J. DPhil1 Author Information 1Department of Mathematics, University of Oxford, Oxford, United Kingdom. 2Iterex Therapeutics, New York, NY. 3Department of Psychiatry, University of Oxford, Oxford, United Kingdom. 4Department of Mathematics, University of Delaware, Newark, DE. Mr. Morrill is supported by the Engineering and Physical Sciences Research Council (EPSRC) under the program grant EP/L015803/1 in collaboration with Iterex Therapeutics. Drs. Kormilitzin and Nevado-Holgado are supported by the Medical Research Council (MRC) under the Pathfinder programme grant MC/PC/17215. Dr. Kormilitzin received support for article research from Wellcome Trust/Charity Open Access Fund. Dr. Swaminathan disclosed government work. Dr. Howison received support for article research from Iterex. Dr. Lyons is supported by the EPSRC under the program grant EP/S026347/1 and by the Alan Turing Institute under the EPSRC grant EP/N510129/1, and his institution received funding from U.K. Research and Innovation MRC. Drs. Kormilitzin, Howison, and Lyons received support for article research fromn Research Councils UK. Mr. Morrill and Drs. Kormilitzin and Lyons are members of the DataSig programme. This work was performed at the Department of Mathematics, University of Oxford, Oxford, United Kingdom. For information regarding this article, E-mail: [email protected] Critical Care Medicine: October 2020 - Volume 48 - Issue 10 - p e976-e981 doi: 10.1097/CCM.0000000000004510 Buy Metrics Abstract Objectives: Patients in an ICU are particularly vulnerable to sepsis. It is therefore important to detect its onset as early as possible. This study focuses on the development and validation of a new signature-based regression model, augmented with a particular choice of the handcrafted features, to identify a patient’s risk of sepsis based on physiologic data streams. The model makes a positive or negative prediction of sepsis for every time interval since admission to the ICU. Design: The data were sourced from the PhysioNet/Computing in Cardiology Challenge 2019 on the “Early Prediction of Sepsis from Clinical Data.” It consisted of ICU patient data from three separate hospital systems. Algorithms were scored against a specially designed utility function that rewards early predictions in the most clinically relevant region around sepsis onset and penalizes late predictions and false positives. Setting: The work was completed as part of the PhysioNet 2019 Challenge alongside 104 other teams. Patients: PhysioNet sourced over 60,000 ICU patients with up to 40 clinical variables for each hour of a patient’s ICU stay. The Sepsis-3 criteria was used to define the onset of sepsis. Interventions: None. Measurements and Main Results: The algorithm yielded a utility function score which was the first placed entry in the official phase of the challenge. Copyright © 2020 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.