High Inter-Patient Variability in Sepsis Evolution: A Hidden Markov Model Analysis. (April 2021)

• Record Type: Journal Article
• Authors: Parente, Jacquelyn D.; Chase, J. Geoffrey; Moeller, Knut; Shaw, Geoffrey M
• Is Part Of: Computer methods and programs in biomedicine. Volume 201(2021)
• Publication Date: 2021-04
• Subjects: Hidden Markov model -- HMM -- classification -- sepsis -- intensive care -- inter-patient variability -- time evolution -- sepsis score; Medicine -- Computer programs -- Periodicals; Biology -- Computer programs -- Periodicals; Computers -- Periodicals; Medicine -- Periodicals; Médecine -- Logiciels -- Périodiques; Biologie -- Logiciels -- Périodiques; Biology -- Computer programs; Medicine -- Computer programs; Periodicals; Electronic journals; 610.28
• DOI: 10.1016/j.cmpb.2021.105956
• Languages: English
• ISSNs: 0169-2607

Abstract

Highlights: A novel physiologically relevant HMM-based classifier for severe sepsis classification is developed and validated. Model validation is performed against clinical data and independent cross validation testing. Validation against statistically relevant metrics. Adding the time evolution of sepsis bio-signals does not significantly improve classifier performance (surprisingly!) Abstract: Background: Severe sepsis and septic shock are common in the intensive care unit (ICU) and contribute significantly to cost and mortality. Early treatment is critical but is confounded by the difficulty of real-time diagnosis. This study uses hidden Markov models (HMMs) to examine whether the time evolution of sepsis can add diagnostic accuracy or value using a proven set of bio-signals. Methods: Clinical data (N=36 patients; 6071 hours), including an hourly personalised insulin sensitivity metric. A two hidden state HMM is created to discriminate diagnosed cases (Severe Sepsis, Septic Shock) from controls (SIRS, Sepsis) states. Diagnostic performance is measured by ROC curves, likelihood ratios (LHRs), sensitivity/specificity, and diagnostic odds-ratios (DOR), for a best-case resubstitution estimate and a worst-case 80/20% repeated holdout analysis. Results: The HMM delivered near perfect results (95% Sensitivity; 96% Specificity) for best-case resubstitution estimates, but was comparatively poor (59% Sensitivity; 61% Specificity) for worst-case repeated holdout estimations. Adding the time evolution of sepsis did not add to the accuracy of diagnosis from using the signals alone without time history. Conclusions: These potentially surprising results indicate significant inter-patient variability in the time evolution of sepsis, preventing effective diagnosis in the context of the bio-signals, data, and HMM topology used. Efforts for improved real-time, early sepsis diagnosis should concentrate on the robustness and efficacy of the bio-signals and data used, as well as the level of model complexity, to create more effective real-time classifiers.