Epidemiology of blood culture-proven bacterial sepsis in children in Switzerland: a population-based cohort study

BACKGROUND
Sepsis is a leading cause of childhood mortality worldwide. We assessed population-based incidence and outcomes of blood culture-proven bacterial sepsis in children in Switzerland.

METHODS
We did a multicentre, prospective, cohort study at ten paediatric hospitals in Switzerland. We included neonates and children younger than 17 years with blood culture-proven bacterial sepsis. Children were eligible if they met criteria for systemic inflammatory response syndrome-according to 2005 paediatric consensus definition- at the time of blood culture sampling. Incidence was calculated by dividing the number of annual sepsis episodes in the study for the years 2012-15 by the end-of-year resident paediatric population in Switzerland. The primary outcome was in-hospital mortality in the first 30 days after sepsis onset.

FINDINGS
Between Sept 1, 2011, and Dec 31, 2015, we enrolled 1096 children to our study. Of 1181 episodes of blood culture-proven bacterial sepsis, 382 (32%) occurred in 379 previously healthy children, 402 (34%) in 391 neonates, and 397 (34%) in 341 children with comorbidities. Incidence was 25·1 cases per 100 000 (95% CI 23·8-26·4) in children and 146·0 cases per 100 000 (133·2-159·6) in neonates. Central line-associated bloodstream infections and primary bloodstream infections accounted for 569 (48%) of 1181 episodes, and organ dysfunction was present in 455 (39%) of 1181 episodes. Escherichia coli (242 of 1181 [20%]), Staphylococcus aureus (177 of 1181 [15%]), coagulase-negative staphylococci (135 of 1181 [11%]), and Streptococcus pneumoniae (118 of 1181 [10%]) were the most prevalent pathogens in our study, accounting for 57% of episodes. The overall case-fatality ratio was 7% (82 of 1181 episodes; 95% CI 5·6-8·6), and it was higher in neonates (11%, 45 of 402 episodes; 8·4-14·8; adjusted odds ratio [OR] 4·41, 95% CI 1·75-11·1) and children with comorbidities (7%, 27 of 397 episodes; 4·6-9·9; OR 4·97, 1·84-13·4) compared with previously healthy children (3%, ten of 382 episodes; 1·3-4·9). The case-fatality ratio was 1% (five of 726 episodes [95% CI 0·3-1·7]) for children without organ dysfunction, which increased to 17% (77 of 455 episodes [13·7-20·8]) when organ dysfunction was present (adjusted OR 4·84, 95% CI 1·40-16·7).

INTERPRETATION
The burden of blood culture-proven bacterial sepsis on child health remains considerable. We recorded key differences in predominant organisms, severity, and outcome between neonates, previously healthy children, and children with comorbidities. Although for most episodes of blood culture-proven bacterial sepsis, no organ dysfunction was seen, presence of organ dysfunction was strongly associated with mortality.