Application of Biomarkers Change to Predict Outcome of Patient With Severe Sepsis

Overview

In 2004, the Surviving Sepsis Campaign (SSC) introduced guidelines for the management of severe sepsis and septic shock, as well as strategies for bedside implementation. The treatment recommendations were organized in two bundles. In an international study, enrolling adult patients with severe sepsis admitted to these intensive care units, investigators found that while mortality from severe sepsis is high (44.5%), compliance with resuscitation and management bundles is generally poor in much of Asia. Investigators need to identify the patients at risk for high in-hospital mortality in order to take appropriate steps. From their past studies, investigators found that sepsis involved inflammation and coagulation. The multiple organ involvement was associated with interaction of novel biomarkers such as cytokines. There is limited data regarding comparing and application of biomarkers of different characteristic on sepsis treatment. A simultaneous detection of multiple cytokines may provide significant prognostic information. For other biomarkers, promising observation data have been put forward, but their potential needs to be evaluated in large-scale, well-designed prospective intervention studies before clinical use can be recommended. Besides many clinical studies on biomarkers were confounded by its lack of standard bundle care for severe sepsis patient. Here investigators performed a systematic study aimed at evaluating 1. the individual and combined diagnostic accuracy of biomarkers for predicting mortality; 2. whether trend change in biomarker level more useful for above prediction; 3. which biomarker or biomarker combination checked can predict patients at risk of evolving with severe organ dysfunctions.

Variables will be tested for their association with the outcome using Pearson chi-square test for categorical data and Mann-Whitney U test for numerical data. Comparison the different groups will be conducted by using Mann-Whitney U test for numerical data and using Pearson chi-square test for categorical data. The time course of biomarker plasma levels will be assessed by analysis of variance. Multivariate analysis will be performed using a logistic regression model to estimate the odds ratio of organ failure and dying, along with the 95% confidence interval (CI). Forward and backward selection procedures will be used to iteratively select the variables potentially related to death. Discrimination will be assessed using the area under the receiver operating characteristic curve to evaluate how well the model distinguished patients who lived from those who died and whether progression of organ dysfunction. A survival analysis will be performed using Kaplan-Meier curves and the log-rank test. Analyses will be completed and a two-tailed p\<0.05 will be considered significant.