Synergistic effects of sepsis and prematurity on neonatal postoperative mortality.
Compared with term neonates, preterm babies are more likely to die from sepsis. However, the combined effects of sepsis and prematurity on neonatal postoperative mortality are largely unknown. Our objective was to quantify the proportion of neonatal postoperative mortality that is attributable to the synergistic effects of preoperative sepsis and prematurity.We performed a multicentre, propensity-score-weighted, retrospective, cohort study of neonates who underwent inpatient surgery across hospitals participating in the United States National Surgical Quality Improvement Program-Pediatric (2012-2017). We assessed the proportion of the observed hazard ratio of mortality and complications that is attributable to the synergistic effect of prematurity and sepsis by estimating the attributable proportion (AP) and its 95% confidence interval (CI).We identified 19 312 neonates who realised a total of 321 321 person-days of postsurgical observations, during which 683 died (mortality rate: 2.1 per 1000 person-days). The proportion of mortality risk that is attributable to the synergistic effect of prematurity and sepsis was 50.5% (AP=50.5%; 95% CI, 28.8-72.3%; P < 0.001). About half of mortality events among preterm neonates with sepsis occurred within 24 h after surgery. Just over 45% of postoperative complications were attributable to the synergistic effect of prematurity and sepsis when both conditions were present (AP=45.8; 95% CI, 13.4-78.1%; P<0.001).Approximately half of postsurgical mortality and complications were attributable to the combined effect of sepsis and prematurity among neonates with both exposures. These neonates typically died within a few days after surgery, indicating a very narrow window of opportunity to predict and prevent mortality.Not applicable.