Drainage as the Initial Surgical Therapy for ELBW Infants with Necrotizing Enterocolitis ( NEe ) or Isolated Intestinal Perforation ( IP ) : Outcomes at 18-22 months Adjusted Age . Short Title : Necrotizing Enterocolitis Surgery Trial ( NEST )

Background. The outcome of ELBW infants with NEC or IP is poor, and the appropriate initial surgical treatment (laparotomy or drainage) is highly controversial. The mortality rate after surgical therapy for ELBW infants approaches 50% and more than half of survivors have been shown to have some element of neurodevelopmental impairment. Until recently, all studies evaluating laparotomy and drainage have focused on death and other outcomes occurring in the immediate post operative period. Three recent prospective studies (the Network observational study2.3 and two randomized trials [Moss and Rees]) found that the death rate in the immediate post operative period does not appear to be significantly influenced by the surgical procedure chosen. Although drainage is widely used and simpler than laparotomy, the Network observational study suggested that laparotomy may improve long term outcome (risk adjusted odds ratio for death or disability (0.55; 95% CI= 0.18 1.67). This observational study is the only prospective study addressing neurodevelopmental outcome according to different surgical management strategies for NEC or IP patients. The Network is best positioned to conduct a trial to resolve this major treatment dilemma, because of an established trial infrastructure and established routine follow up programs. An editorial in the New England Journal of Medicine pointed out the limitations of current studies evaluating surgical therapies in NEC and highlighted the importance of measuring neurodevelopmental outcome beyond nursery discharge. Over the past several years, each of the above-referenced studies have been very valuable in providing different pieces of evidence regarding which surgical therapy should be considered most effective in ELBW infants with NEC or isolated intestinal perforation. The trial proposed herein is a natural extension of these studies and will provide critically needed information. Hypotheses. The primary hypothesis is that survival without neurodevelopmental impairment (NOI) at 18­ 22 months adjusted age will be greater with initial laparotomy rather than with initial drainage. Methods. Inclusion criteria are 1) birth weight of ~ 1,000 g, 2) a decision to perform surgery for suspected NEC or IP, 3) the infant is less than or equal to 8 weeks of age (8 017 weeks or less) at the time of eligibility assessment, and 4) patient is at a center able to perform both laparotomy and drainage. Exclusion criteria are: 1) major anomaly which influences likelihood of developing primary outcome or affects surgical treatment considerations; 2) congenital infection; 3) prior NEC or IP; 4) prior laparotomy or peritoneal drain placement; and 5) follow-up unlikely and 6) infant for whom full support is not provided (including surgical treatment). To facilitate enrollment, the trial will be discussed with the parents of potentially eligible infants when NEC or IP is first suspected. Consent will be diligently sought for randomization utilizing members of the multidisciplinary research team (i.e. research coordinator, neonatologist, pediatric surgeon). For infants that are not randomized (due to refusal of consent or unwillingness of the neonatologist andlor surgeon to enroll the infant), consent will be sought to enroll the infant into the study as a member of the "preference cohort". These infants will be treated according to the preference of the surgeon and neonatologist and will have similar data collection as for randomized infants. Enrollment in the preference cohort was terminated on February 14, 2013. Infants will be enrolled by calling RTI International as soon as a decision to perform surgery is made (or via a computer randomization process). Randomized infants will be stratified according to two variables: center and according to the overall risk for death or NOI (higher I lower). The risk stratification formula was primarily developed in the prospective observational model and includes the following variables: birth-weight, gestational age, vasopressor requirement (yes I no), if infant is on high frequency (oscillating or jet) ventilation or not (any modality of conventional ventilation or no mechanical ventilation), pH, Fi02, and preoperative diagnosis (NEC or IP). Treatment beyond the initial surgical management will be unaffected by the trial. Oata collection (beyond that for the generic data base) will be limited, and the forms will be based on those used in the observational study. Outcome at 18-22 months will be assessed by evaluators who are masked to the details of the operative intervention(s) that have been performed using standard Network assessments. Analysis & Interpretation. The primary outcome (death or NOI at 18-22 months corrected age) will be analyzed in standard logistic regression equations in which the predictor variables include center, and level of risk at enrollment as determined by the method described above. Separate analyses will be used to assess whether there is an interaction between treatment (laparotomy or drainage) and disease (NEC or IP). Secondary analyses of randomized patients will be performed using Bayesian methods to assess the (posterior) probability that laparotomy reduces death or NOI relative to that with drainage. This document is provided for reference purposes only. Persons with disabilities having difficulty accessing information in this document should e-mail NICHD FOIA Office at NICHDFOIARequest@mail.nih.gov for assistance.