The association of coagulation indicators with in-hospital mortality and 1-year mortality of patients with sepsis at ICU admissions: A retrospective cohort study.

ABSTRACT

BACKGROUND: Coagulation activation is the host's response to pathogens during sepsis and is considered to be one of the reasons for tissue damage and multiple organ failure. This study is designed to evaluate whether the alterations of coagulation indicators are related to in-hospital mortality and 1-year mortality of patients with sepsis. METHOD: Data of all 2258 patients were extracted from the database Multiparameter Intelligent Monitoring in Intensive Care III (MIMIC III). The relationship between the in-hospital mortality of patients with sepsis and coagulation indicators was analyzed with a receiver operating characteristic (ROC) curve analysis and logistic regression model. Effects of coagulation indicators on patients' 1-year mortality were determined by using a Cox hazard regression model, and clinical experience or quintiles were used to classify the activated partial thromboplastin time (APTT) to determine the cutoff values to explore segmentation effects. RESULT: International normalized ratio (INR) was positively associated with hospital mortality of patients with sepsis after adjusting confounders with an odds ratio (OR) of 1.86 [95% confidence interval (CI), 1.37-2.52], and a hazard ratio (HR) of 1.465[95%CI(1.24-1.74)] for 1-year mortality, respectively. 1-year mortality of patients with sepsis demonstrated a U-shaped relationship with APTT, ranging from 25 to 37, indicating the lowest risk. The adjusted HR (95% CI) values for 1-year mortality of septic patients with risk values <25 and >37 were 1.493 (1.02, 2.19) and 1.379 (1.06, 1.79), respectively. CONCLUSION: Increased INR in critically ill septic patients is related to greater in-hospital mortality and 1-year mortality. A U-shaped relationship was found between APTT and 1-year mortality of patients with sepsis.