[Clinical risk factors for capillary leak syndrome in children with sepsis]

To investigate the clinical features of capillary leak syndrome (CLS) in children with sepsis, and to analyze its risk factors.

Clinical data of 384 children with sepsis was studied retrospectively. They included 304 cases of general sepsis, 54 cases of severe sepsis and 26 cases of septic shock, and were divided into non-CLS (n=356) and CLS groups (n=28). Univariate analysis was performed for each of the following variables: sex, age, malnutrition, anemia, coagulation disorders, white blood cell count, C-reactive protein (CRP), procalcitonin (PCT), tumor necrosis factor (TNF), interleukin (IL)-1, IL-6, blood glucose, lactic acid, Pediatric Risk of Mortality (PRISM) III score, pediatric critical illness score (PICS), severe sepsis and number of failed organs≥3. The statistically significant variables (as independent variables) were subjected to multivariate logistic regression analysis.

The incidence rate of CLS in children with septic shock, severe sepsis and general sepsis were 42.3%, 20.1% and 1.3%, respectively, with significant differences among them (P<0.01). There were significant differences in anemia, coagulation disorders, CRP, PCT>2 ng/mL, TNF, IL-1, IL-6, blood glucose, lactic acid, PRISM III score, PICS and number of failed organs≥3 between the non-CLS and CLS groups (P<0.05). Severe sepsis/shock and PRISM III score were the independent risk factors for CLS in children with sepsis.

The severity of sepsis and PRISM III score are positively correlated with the incidence of CLS in children with sepsis. Early monitoring of such factors as infection markers and blood glucose in children with severe sepsis and high PRISM III score may contribute to early diagnosis and effective intervention, thus reducing the mortality from CLS in children with sepsis.