Ritankar Das

A computational approach to early sepsis detection

OBJECTIVE To develop high-performance early sepsis prediction technology for the general patient population. METHODS Retrospective analysis of adult patients admitted to the intensive care unit (from the MIMIC II dataset) who were not septic at the time of admission. RESULTS A sepsis early warning algorithm, InSight, was developed and applied to the prediction of sepsis up to three hours prior to a patients first five hour Systemic Inflammatory Response Syndrome (SIRS) episode. When applied to a never-before-seen set of test patients, InSight predictions demonstrated a sensitivity of 0.90 (95% CI: 0.89-0.91) and a specificity of 0.81 (95% CI: 0.80-0.82), exceeding or rivaling that of existing biomarker detection methods. Across predictive times up to three hours before a sustained SIRS event, InSight maintained an average area under the ROC curve of 0.83 (95% CI: 0.80-0.86). Analysis of patient sepsis risk showed that contributions from the coevolution of multiple risk factors were more important than the contributions from isolated individual risk factors when making predictions further in advance. CONCLUSIONS Sepsis can be predicted at least three hours in advance of onset of the first five hour SIRS episode, using only nine commonly available vital signs, with better performance than methods in standard practice today. High-order correlations of vital sign measurements are key to this prediction, which improves the likelihood of early identification of at-risk patients.

Discharge recommendation based on a novel technique of homeostatic analysis

Objective: We propose a computational framework for integrating diverse patient measurements into an aggregate health score and applying it to patient stability prediction. Materials and Methods: We mapped retrospective patient data from the Multiparameter Intelligent Monitoring in Intensive Care (MIMIC) II clinical database into a discrete multidimensional space, which was searched for measurement combinations and trends relevant to patient outcomes of interest. Patient trajectories through this space were then used to make outcome predictions. As a case study, we built AutoTriage, a patient stability prediction tool to be used for discharge recommendation. Results: AutoTriage correctly identified 3 times as many stabilizing patients as existing tools and achieved an accuracy of 92.9% (95% CI: 91.6–93.9%), while maintaining 94.5% specificity. Analysis of AutoTriage parameters revealed that interdependencies between risk factors comprised the majority of each patient stability score. Discussion: AutoTriage demonstrated an improvement in the sensitivity of existing stability prediction tools, while considering patient safety upon discharge. The relative contributions of risk factors indicated that time-series trends and measurement interdependencies are most important to stability prediction. Conclusion: Our results motivate the application of multidimensional analysis to other clinical problems and highlight the importance of risk factor trends and interdependencies in outcome prediction.

High-performance detection and early prediction of septic shock for alcohol-use disorder patients

Background The presence of Alcohol Use Disorder (AUD) complicates the medical conditions of patients and increases the difficulty of detecting and predicting the onset of septic shock for patients in the ICU. Methods We have developed a high-performance sepsis prediction algorithm, InSight, which outperforms existing methods for AUD patient populations. InSight analyses a combination of singlets, doublets, and triplets of clinical measurements over time to generate a septic shock risk score. AUD patients obtained from the MIMIC III database were used in this retrospective study to train InSight and compare performance with the Modified Early Warning Score (MEWS), the Simplified Acute Physiology Score (SAPS II), and the Systemic Inflammatory Response Syndrome (SIRS) for septic shock prediction and detection. Results From 4-fold cross validation, InSight performs particularly well on diagnostic odds ratio and demonstrates a relatively high Area Under the Receiver Operating Characteristic (AUROC) metric. Four hours prior to onset, InSight had an average AUROC of 0.815, and at the time of onset, InSight had an average AUROC value of 0.965. When applied to patient populations where AUD may complicate prediction methods of sepsis, InSight outperforms existing diagnostic tools. Conclusions Analysis of the higher order correlations and trends between relevant clinical measurements using the InSight algorithm leads to more accurate detection and prediction of septic shock, even in cases where diagnosis may be confounded by AUD.

Energy landscapes for a machine learning application to series data

Methods developed to explore and characterise potential energy landscapes are applied to the corresponding landscapes obtained from optimisation of a cost function in machine learning. We consider neural network predictions for the outcome of local geometry optimisation in a triatomic cluster, where four distinct local minima exist. The accuracy of the predictions is compared for fits using data from single and multiple points in the series of atomic configurations resulting from local geometry optimisation and for alternative neural networks. The machine learning solution landscapes are visualised using disconnectivity graphs, and signatures in the effective heat capacity are analysed in terms of distributions of local minima and their properties.

Using electronic health record collected clinical variables to predict medical intensive care unit mortality

Background Clinical decision support systems are used to help predict patient stability and mortality in the Intensive Care Unit (ICU). Accurate patient information can assist clinicians with patient management and in allocating finite resources. However, systems currently in common use have limited predictive value in the clinical setting. The increasing availability of Electronic Health Records (EHR) provides an opportunity to use medical information for more accurate patient stability and mortality prediction in the ICU. Objective Develop and evaluate an algorithm which more accurately predicts patient mortality in the ICU, using the correlations between widely available clinical variables from the EHR. Methods We have developed an algorithm, AutoTriage, which uses eight common clinical variables from the EHR to assign patient mortality risk scores. Each clinical variable produces a subscore, and combinations of two or three discretized clinical variables also produce subscores. A combination of weighted subscores produces the overall score. We validated the performance of this algorithm in a retrospective study on the MIMIC III medical ICU dataset. Results AutoTriage 12 h mortality prediction yields an Area Under Receiver Operating Characteristic value of 0.88 (95% confidence interval 0.86 to 0.88). At a sensitivity of 80%, AutoTriage maintains a specificity of 81% with a diagnostic odds ratio of 16.26. Conclusions Through the multidimensional analysis of the correlations between eight common clinical variables, AutoTriage provides an improvement in the specificity and sensitivity of patient mortality prediction over existing prediction methods.

A computational approach to mortality prediction of alcohol use disorder inpatients

BACKGROUND Health information technologies can assist clinicians in the Intensive Care Unit (ICU) by providing additional analysis of patient stability. However, because patient diagnoses can be confounded by chronic alcohol use, the predictive value of existing systems is suboptimal. Through the use of Electronic Health Records (EHR), we have developed computer software called AutoTriage to generate accurate predictions through multi-dimensional analysis of clinical variables. We analyze the performance of AutoTriage on the Alcohol Use Disorder (AUD) subpopulation in this study, and build on results we reported for AutoTriage performance on the general population in previous work. METHODS AUD-related ICD-9 codes were used to obtain a patient population from MIMIC III ICU dataset for a retrospective study. Patient mortality risk score is generated through analysis of eight EHR-based clinical variables. The score is determined by combining weighted subscores, each of which are obtained from singlets, doublets or triplets of one or more of the eight continuous-valued clinical variable inputs. A temporally updating risk score is computed with a continuously revised 12-hour mortality prediction. RESULTS Among AUD patients, in a non-overlapping test set, AutoTriage outperforms existing systems with an Area Under Receiver Operating Characteristic (AUROC) value of 0.934 for 12-h mortality prediction. At a sensitivity of 90%, AutoTriage achieves a specificity of 80%, positive predictive value of 40%, negative predictive value of 89%, and an Odds Ratio of 36. CONCLUSIONS For mortality prediction, AutoTriage demonstrates improvements in both the accuracy and the Odds Ratio over current systems among the AUD patient population.

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

Objectives We validate a machine learning-based sepsis-prediction algorithm (InSight) for the detection and prediction of three sepsis-related gold standards, using only six vital signs. We evaluate robustness to missing data, customisation to site-specific data using transfer learning and generalisability to new settings. Design A machine-learning algorithm with gradient tree boosting. Features for prediction were created from combinations of six vital sign measurements and their changes over time. Setting A mixed-ward retrospective dataset from the University of California, San Francisco (UCSF) Medical Center (San Francisco, California, USA) as the primary source, an intensive care unit dataset from the Beth Israel Deaconess Medical Center (Boston, Massachusetts, USA) as a transfer-learning source and four additional institutions’ datasets to evaluate generalisability. Participants 684 443 total encounters, with 90 353 encounters from June 2011 to March 2016 at UCSF. Interventions None. Primary and secondary outcome measures Area under the receiver operating characteristic (AUROC) curve for detection and prediction of sepsis, severe sepsis and septic shock. Results For detection of sepsis and severe sepsis, InSight achieves an AUROC curve of 0.92 (95% CI 0.90 to 0.93) and 0.87 (95% CI 0.86 to 0.88), respectively. Four hours before onset, InSight predicts septic shock with an AUROC of 0.96 (95% CI 0.94 to 0.98) and severe sepsis with an AUROC of 0.85 (95% CI 0.79 to 0.91). Conclusions InSight outperforms existing sepsis scoring systems in identifying and predicting sepsis, severe sepsis and septic shock. This is the first sepsis screening system to exceed an AUROC of 0.90 using only vital sign inputs. InSight is robust to missing data, can be customised to novel hospital data using a small fraction of site data and retains strong discrimination across all institutions.

Using Transfer Learning for Improved Mortality Prediction in a Data-Scarce Hospital Setting

Algorithm–based clinical decision support (CDS) systems associate patient-derived health data with outcomes of interest, such as in-hospital mortality. However, the quality of such associations often depends on the availability of site-specific training data. Without sufficient quantities of data, the underlying statistical apparatus cannot differentiate useful patterns from noise and, as a result, may underperform. This initial training data burden limits the widespread, out-of-the-box, use of machine learning–based risk scoring systems. In this study, we implement a statistical transfer learning technique, which uses a large “source” data set to drastically reduce the amount of data needed to perform well on a “target” site for which training data are scarce. We test this transfer technique with AutoTriage, a mortality prediction algorithm, on patient charts from the Beth Israel Deaconess Medical Center (the source) and a population of 48 249 adult inpatients from University of California San Francisco Medical Center (the target institution). We find that the amount of training data required to surpass 0.80 area under the receiver operating characteristic (AUROC) on the target set decreases from more than 4000 patients to fewer than 220. This performance is superior to the Modified Early Warning Score (AUROC: 0.76) and corresponds to a decrease in clinical data collection time from approximately 6 months to less than 10 days. Our results highlight the usefulness of transfer learning in the specialization of CDS systems to new hospital sites, without requiring expensive and time-consuming data collection efforts.

Effect of a machine learning-based severe sepsis prediction algorithm on patient survival and hospital length of stay: a randomised clinical trial

Introduction Several methods have been developed to electronically monitor patients for severe sepsis, but few provide predictive capabilities to enable early intervention; furthermore, no severe sepsis prediction systems have been previously validated in a randomised study. We tested the use of a machine learning-based severe sepsis prediction system for reductions in average length of stay and in-hospital mortality rate. Methods We conducted a randomised controlled clinical trial at two medical-surgical intensive care units at the University of California, San Francisco Medical Center, evaluating the primary outcome of average length of stay, and secondary outcome of in-hospital mortality rate from December 2016 to February 2017. Adult patients (18+) admitted to participating units were eligible for this factorial, open-label study. Enrolled patients were assigned to a trial arm by a random allocation sequence. In the control group, only the current severe sepsis detector was used; in the experimental group, the machine learning algorithm (MLA) was also used. On receiving an alert, the care team evaluated the patient and initiated the severe sepsis bundle, if appropriate. Although participants were randomly assigned to a trial arm, group assignments were automatically revealed for any patients who received MLA alerts. Results Outcomes from 75 patients in the control and 67 patients in the experimental group were analysed. Average length of stay decreased from 13.0 days in the control to 10.3 days in the experimental group (p=0.042). In-hospital mortality decreased by 12.4 percentage points when using the MLA (p=0.018), a relative reduction of 58.0%. No adverse events were reported during this trial. Conclusion The MLA was associated with improved patient outcomes. This is the first randomised controlled trial of a sepsis surveillance system to demonstrate statistically significant differences in length of stay and in-hospital mortality. Trial registration NCT03015454.

Pediatric Severe Sepsis Prediction Using Machine Learning

Early detection of pediatric severe sepsis is necessary in order to administer effective treatment. In this study, we assessed the efficacy of a machine-learning-based prediction algorithm applied to electronic healthcare record (EHR) data for the prediction of severe sepsis onset. The resulting prediction performance was compared with the Pediatric Logistic Organ Dysfunction score (PELOD-2) and pediatric Systemic Inflammatory Response Syndrome score (SIRS) using cross-validation and pairwise t-tests. EHR data were collected from a retrospective set of de-identified pediatric inpatient and emergency encounters drawn from the University of California San Francisco (UCSF) Medical Center, with encounter dates between June 2011 and March 2016. Patients (n = 11,127) were 2-17 years of age and 103 [0.93%] were labeled severely septic. In four-fold cross-validation evaluations, the machine learning algorithm achieved an AUROC of 0.912 for discrimination between severely septic and control pediatric patients at onset and AUROC of 0.727 four hours before onset. Under the same measure, the prediction algorithm also significantly outperformed PELOD-2 (p < 0.05) and SIRS (p < 0.05) in the prediction of severe sepsis four hours before onset. This machine learning algorithm has the potential to deliver high-performance severe sepsis detection and prediction for pediatric inpatients.