LIV-4: A novel model for predicting transplant-free survival in critically ill cirrhotics

Abstract

BACKGROUND Critically ill patients with cirrhosis, particularly those with acute decompensation, have higher mortality rates in the intensive care unit (ICU) than patients without chronic liver disease. Prognostication of short-term mortality is important in order to identify patients at highest risk of death. None of the currently available prognostic models have been widely accepted for use in cirrhotic patients in the ICU, perhaps due to complexity of calculation, or lack of universal variables readily available for these patients. We believe a survival model meeting these requirements can be developed, to guide therapeutic decision-making and contribute to cost-effective healthcare resource utilization. AIM To identify markers that best identify likelihood of survival and to determine the performance of existing survival models. METHODS Consecutive cirrhotic patients admitted to a United States quaternary care center ICU between 2008-2014 were included and comprised the training cohort. Demographic data and clinical laboratory test collected on admission to ICU were analyzed. Area under the curve receiver operator characteristics (AUROC) analysis was performed to assess the value of various scores in predicting in-hospital mortality. A new predictive model, the LIV-4 score, was developed using logistic regression analysis and validated in a cohort of patients admitted to the same institution between 2015-2017. RESULTS Of 436 patients, 119 (27.3%) died in the hospital. In multivariate analysis, a combination of the natural logarithm of the bilirubin, prothrombin time, white blood cell count, and mean arterial pressure was found to most accurately predict in-hospital mortality. Derived from the regression coefficients of the independent variables, a novel model to predict inpatient mortality was developed (the LIV-4 score) and performed with an AUROC of 0.86, compared to the Model for End-Stage Liver Disease, Chronic Liver Failure-Sequential Organ Failure Assessment, and Royal Free Hospital Score, which performed with AUROCs of 0.81, 0.80, and 0.77, respectively. Patients in the internal validation cohort were substantially sicker, as evidenced by higher Model for End-Stage Liver Disease, Model for End-Stage Liver Disease-Sodium, Acute Physiology and Chronic Health Evaluation III, SOFA and LIV-4 scores. Despite these differences, the LIV-4 score remained significantly higher in subjects who expired during the hospital stay and exhibited good prognostic values in the validation cohort with an AUROC of 0.80. CONCLUSION LIV-4, a validated model for predicting mortality in cirrhotic patients on admission to the ICU, performs better than alternative liver and ICU-specific survival scores.

State of the Art of Machine Learning–Enabled Clinical Decision Support in Intensive Care Units: Literature Review

“…Patients with sepsis and infection remain one of the most studied populations in terms of mortality (generally 28 days) [68][69][70][71][72], followed by acute kidney injury [72][73][74][75]. There is an increasing trend in outcome prediction studies in critically ill patients with liver disease-acute liver injury [76,77], cirrhosis [77], and advanced liver disease [78] have been studied using machine learning. In patients with severe cancer, 30- [79] and 120-day [80] survival rates were studied retrospectively with logistic regression models.…”