Risk factors of in-hospital mortality in critically ill patients with intracerebral hemorrhage

Objective To analyze the risk factors of in-hospital mortality in patients with intracerebral hemorrhage (ICH) in the intensive care unit. Methods Patients with intracerebral hemorrhage were retrospectively collected from January 2013 to January 2018 in the Department of Intensive Care Unit, Zhongnan Hospital of Wuhan University. Patients were excluded aged less than 18 years, pregnant women, the onset time of more than 7 days, the length of hospital stay of less than 48 hours, lack of renal function outcomes within 24 hours after admission, the glomerular filtration rate (eGFR) of lower than 15 mL/(min·1.73 m2), the history of chronic kidney disease, regular dialysis and renal transplantation, and incomplete data. Clinical data were collected from baseline characteristics, past history, and laboratory examination. The included patients were divided into the in-hospital non-survival group and the survival group. SPSS 20.0 software as used for statistical analysis, the binary Logistic regression analysis was performed to evaluate risk factors of in-hospital mortality with intracerebral hemorrhage, the prognosis was assessed by receiver operating characteristic (ROC) curve and survival curve (Kaplan-Meier). A P<0.05 was considered statistically significant. Results In this single-center retrospective study, a total of 300 patients were enrolled, including 96 patients in the hospital non-survival group and 204 patients in the survival group. The incidence of in-hospital death in patients with intracerebral hemorrhage in ICU was 32%. Multivariate analysis demonstrated that the risk factors of in-hospital mortality were lower GCS score (OR=0.629, 95%CI: 0.523-0.757, P<0.01), higher APACHEⅡ score (OR=1.590, 95%CI: 1.369-1.847, P<0.01), elevated leukocytes (OR=1.082, 95%CI: 1.028-1.139, P=0.002) and the incidence of acute kidney injury (AKI) (OR=6.978, 95%CI: 3.381-14.405, P<0.01). The ROC curve demonstrated that the area under curve (AUC) of APACHEⅡ score was the largest with a sensitivity and specificity of 73.96% and 75.98%, respectively, which can better predict the mortality of patients with cerebral hemorrhage. Kaplan-Meier survival curve showed that in-hospital survival rate of non-AKI patients were higher than that of AKI patients (P<0.01). Conclusions Lower GCS score, higher APACHEⅡ score, elevated white blood cells and AKI are risk factors for predicting in-hospital mortality in patients with intracerebral hemorrhage in the ICU. Therefore, early identification and treatment should be adopted in these high-risk populations. Key words: Intracerebral hemorrhage; In-hospital mortality; Intensive care unit; Risk factors