Plasma and Urinary Biomarkers Improve Prediction of Mortality through 1 Year in Intensive Care Patients: An Analysis from FROG-ICU.

Abstract

This study aimed to assess the value of blood and urine biomarkers in addition to routine clinical variables in risk stratification of patients admitted to ICU. Multivariable prognostic models were developed in this post hoc analysis of the French and EuRopean Outcome ReGistry in Intensive Care Units study, a prospective observational study of patients admitted to ICUs. The study included 2087 patients consecutively admitted to the ICU who required invasive mechanical ventilation or a vasoactive agent for more than 24 h. The main outcome measures were in-ICU, in-hospital, and 1 year mortality. Models including only SAPS II or APACHE II scores had c-indexes for in-hospital and 1 year mortality of 0.64 and 0.65, and 0.63 and 0.61, respectively. The c-indexes for a model including age and estimated glomerular filtration rate were higher at 0.69 and 0.67, respectively. Models utilizing available clinical variables increased the c-index for in-hospital and 1 year mortality to 0.80 and 0.76, respectively. The addition of biomarkers and urine proteomic markers increased c-indexes to 0.83 and 0.78. The commonly used scores for risk stratification in ICU patients did not perform well in this study. Models including clinical variables and biomarkers had significantly higher predictive values.