Morbidity Associated with Laparotomy-confirmed Spontaneous Intestinal Perforation: A Prospective Multicenter Analysis

Background: Spontaneous intestinal perforation (SIP) and necrotizing enterocolitis (NEC) are the most common gastrointestinal emergencies affecting preterm infants. They are considered distinct pathologic entities, with the mortality of NEC known to be significantly higher than that of SIP. However, differences in morbidity between these two diseases have not been well-studied. The aim of this study is to quantify risk-adjusted length of stay (LOS) and morbidities at discharge among very low birth weight (VLBW) survivors of SIP and NEC.

Methods: An analysis of prospectively collected data regarding VLBW infants (birth weight <1500g or gestational age 22-29 weeks) from 790 North American centers over a 5-year period (2015-2019) was performed. All included infants had laparotomy confirmation of SIP or NEC. The primary outcome was length of stay (LOS) among survivors. A negative binomial regression model was used to calculate adjusted length of stay, controlling for gestational age, small for gestational age, sex, inborn/outborn status, multiple birth, and clustering of infants within hospitals. Secondary outcomes included growth morbidity (weight <10th percentile) and need for supplemental oxygen, both at discharge .

Results: In this study of 193,048 VLBW infants, 1523 had laparotomy-confirmed SIP and 2601 had laparotomy-confirmed NEC . Infants with SIP were more likely than infants with NEC to be <25 weeks gestational age at birth (44.8% vs 33.1%) and more frequently required resuscitation in the delivery room in the form of intubation (74.9% vs 67.9%) or cardiac compressions (8.1% vs 6.6%). Infants with SIP were also more likely to have: PDA ligation (16.9% vs 12.9%), severe intraventricular hemorrhage (25.0% vs 19.0%), and tracheostomy (2.9% vs 1.5%). The adjusted LOS was similar for infants with SIP (92 days, 95% CI 88-95) and infants with NEC (88 days, 95% CI 85-91), but significantly higher for both when compared to infants without SIP or NEC (68 days, 95% CI 67-69; Figure 1). The incidence of growth morbidity at discharge was similar between SIP and NEC (74.2% vs 75.3 %), but higher for both when compared to infants without SIP or NEC (47.7%). Infants with SIP were more likely to require supplemental oxygen at discharge (34.9%) than infants with NEC (24.4%) and infants without SIP or NEC (13.6%; Figure 2).

Conclusion: VLBW infants with laparotomy-confirmed SIP have an adjusted length of stay of nearly 3 months, similar to infants with laparotomy-confirmed NEC but significantly longer than infants without SIP or NEC. The high incidence of growth morbidity and frequent need for supplemental oxygen at discharge are also notable among infants with SIP. These findings demonstrate the high morbidity associated with laparotomy-confirmed SIP, helping surgeons counsel families while also providing clear targets for improvement of care in these complex patients.