Creating an automated trigger for sepsis clinical decision support at emergency department triage using machine learning

Abstract

ObjectiveTo demonstrate the incremental benefit of using free text data in addition to vital sign and demographic data to identify patients with suspected infection in the emergency department.MethodsThis was a retrospective, observational cohort study performed at a tertiary academic teaching hospital. All consecutive ED patient visits between 12/17/08 and 2/17/13 were included. No patients were excluded. The primary outcome measure was infection diagnosed in the emergency department defined as a patient having an infection related ED ICD-9-CM discharge diagnosis. Patients were randomly allocated to train (64%), validate (20%), and test (16%) data sets. After preprocessing the free text using bigram and negation detection, we built four models to predict infection, incrementally adding vital signs, chief complaint, and free text nursing assessment. We used two different methods to represent free text: a bag of words model and a topic model. We then used a support vector machine to build the prediction model. We calculated the area under the receiver operating characteristic curve to compare the discriminatory power of each model.ResultsA total of 230,936 patient visits were included in the study. Approximately 14% of patients had the primary outcome of diagnosed infection. The area under the ROC curve (AUC) for the vitals model, which used only vital signs and demographic data, was 0.67 for the training data set, 0.67 for the validation data set, and 0.67 (95% CI 0.65–0.69) for the test data set. The AUC for the chief complaint model which also included demographic and vital sign data was 0.84 for the training data set, 0.83 for the validation data set, and 0.83 (95% CI 0.81–0.84) for the test data set. The best performing methods made use of all of the free text. In particular, the AUC for the bag-of-words model was 0.89 for training data set, 0.86 for the validation data set, and 0.86 (95% CI 0.85–0.87) for the test data set. The AUC for the topic model was 0.86 for the training data set, 0.86 for the validation data set, and 0.85 (95% CI 0.84–0.86) for the test data set.ConclusionCompared to previous work that only used structured data such as vital signs and demographic information, utilizing free text drastically improves the discriminatory ability (increase in AUC from 0.67 to 0.86) of identifying infection.

Early detection of sepsis utilizing deep learning on electronic health record event sequences

“…The World Health Organization estimates that more than six million people die of sepsis annually, and many of these deaths are preventable [2]. In the United States, severe sepsis Previous studies have shown that machine learning (ML) models trained from data in individual patient electronic health records (EHR) may be used for the early detection of sepsis [7,8,9,10,11,12]. The ML models for sepsis detection far exceed the predictive ability of existing clinical early warning system scores, such as the National Early Warning Score (NEWS) [7,9,10,12,13,14].…”

Early prediction of circulatory failure in the intensive care unit using machine learning

“…Machine learning techniques excel in the analysis of complex signals in data-rich environments 11,12 . Recent advances in deep learning, reinforcement learning, and other machine learning techniques have enabled and popularized their use in mathematics 13 , engineering 14 , biology 15 and medicine [16][17][18][19][20][21][22][23][24][25][26]77 . The abundance of data collected in the ICU and the importance of timely decision-making are key to the growing interest of using of machine learning in this setting.…”

Multicenter validation of a sepsis prediction algorithm using only vital sign data in the emergency department, general ward and ICU

“…It is important to note that we designed the study as a classification task rather than a time-to-event modeling experiment, because the former is significantly more common in the literature [28][29][30][31]. The alternative would not allow for the use of an established, standard set of performance metrics such as AUROC and specificity without custom modification, and would make it more difficult to compare the present study to prior work in the field.…”