Mortality after emergency abdominal operations in premature infants

Context/background: To determine risk of 30-day mortality for premature infants undergoing abdominal operations during the first 2 months of life and to identify risk factors for perioperative mortality using available demographic and clinical variables of interest. Basic procedures: Retrospective descriptive analysis of premature infants (gestational age less than or equal to 36 weeks) undergoing abdominal operations during the first 2 months of life using the American College of Surgeon's National Surgical Quality Improvement Project Pediatric (NSQIP-P, 2012–2015) database. A stepwise logistic regression model incorporating multiple demographic and clinical factors was constructed to identify independent predictors of 30-day mortality. Findings: A total of 1554 premature infants were identified who underwent abdominal operations during the first 2 months of life. Unadjusted 30-day mortality ranged from 31% for infants born < 24 weeks gestational age to 4.9% for those born at 35–36 weeks. Increased gestational age corresponded to decreased risk of mortality but week-by-week was not independently predictive of mortality in multivariate modeling. Female sex (aOR 1.51, 95% C.I. 1.08–2.10, p = 0.014), inotrope support (aOR 3.46, 95% C.I. 2.43–4.92, p < 0.001), ventilator use (aOR 2.86, 95% C.I. 1.56–5.25, p < 0.001) and American Society of Anesthesiologists (ASA) class 3 (aOR 4.14, 95% C.I. 1.58–10.81, p = 0.004) at time of operation were all associated with significantly increased risk of 30-day mortality. On stepwise logistic regression incorporating only those variables with statistical significance, female sex, inotrope, and ventilator support retained statistical significance. Conclusions: Premature infants undergoing abdominal operations during the first 2 months of life have expectedly high risk of 30-day mortality. Female sex, inotrope, and ventilator support are independently associated with increased risk of mortality and can be incorporated into a model where, if present, risk of mortality is greater than 14.2%.