Epidemiological investigation and logistic regression analysis of nosocomial fungal infection in severe sepsis patients

Abstract

Objective To investigate the epidemiological characteristics of nosocomial fungal infections in severe sepsis patients and analyze the influencing factors by logistic regression. Methods The clinical data of 578 patients with severe sepsis admitted to our hospital from January 2008 to January 2016 were retrospectively analyzed. The epidemiological characteristics were investigated. The possible influencing factors of nosocomial fungal infection were analyzed by univariate analysis and multivariate logistic regression analysis. Results Among 578 cases of severe sepsis, 215 cases were accompanied by nosocomial fungal infection (infection rate 37.20%). The infection sites were mostly lungs and the fungal types were Candida albicans. The prognosis of the patients was poor. The survival time was less than 90 days, accounting for 42.33%(91/215). There were significant differences in age, length of stay in intensive care unit (ICU), duration of more than two antibiotics, use of ventilators, hormones, invasive procedures, acute physiology and chronic health evaluation Ⅱ (APACHE Ⅱ) score between the infected group and the uninfected group (P 7 d (OR=3.287), use steroid (OR=2.141). Conclusions Severe sepsis with hospital acquired fungal infection is more common in the lung, and most of them are Candida albicans infection, with poor prognosis and high mortality. ICU stay ≥100 h, use respirator and steroid, invasive operation, age ≥65 years, and continuous use of two antibiotics >7d are risk factors. Specific preventive measures should be taken to reduce fungal infections. Key words: Sepsis; Mycoses/EP; Cross infection; Hospitals; Factor analysis, statistical