Clinical predictors for severe sepsis in patients with necrotizing fasciitis: an observational cohort study in northern Thailand

BACKGROUND: Necrotizing fasciitis (NF) is a life-threatening infection of skin and fascia. Its progress is extremely fast, with extensive necrosis. Delay in treatment, with subsequent huge soft tissue loss and associated severe sepsis, remains a major cause of death in the management of NF. OBJECTIVE: The aim of this study was to explore clinical characteristics that may be used to predict severe sepsis in patients with NF, in the context of routine clinical practice in northern Thailand. METHODS: A retrospective observational cohort study was conducted. The patient cohort in this study consisted of all patients who were diagnosed with NF by surgical or pathological confirmation. The follow-up period started with the admission date and ended with the discharge date. The clinical variables were collected from patients registered at three provincial hospitals in northern Thailand from 2009 to 2012. The clinical predictors for severe sepsis were analyzed using multivariable risk regression. RESULTS: A total of 1,452 patients were diagnosed with NF, either with severe sepsis (n=237 [16.3%]) or without severe sepsis (n=1,215 [83.7%]). From the multivariable analysis, female sex (relative risk [RR] =1.51; 95% confidence interval [CI] =1.04–2.20), diabetes mellitus (RR =1.40; 95% CI =1.25–1.58), chronic heart disease (RR =1.31; 95% CI =1.15–1.49), hemorrhagic bleb (RR =1.47; 95% CI =1.32–1.63), skin necrosis (RR =1.45; 95% CI =1.34–1.57), and serum protein <6 g/dL (RR =2.67; 95% CI =1.60–4.47) were all predictive factors for severe sepsis. CONCLUSION: The clinical predictors for severe sepsis in patients with suspicion of NF included female sex, diabetes mellitus, chronic heart disease, hemorrhagic bleb, skin necrosis, and serum protein <6 d/dL. The risk ratio was much higher in patients with total protein less than 6 g/dL, which is associated with malnutrition. Therefore, provision of sufficient nutritional support and close monitoring for these clinical predictors may be beneficial to reduce morbidity and mortality.