Intestinal atresia is an important cause of neonatal intestinal obstruction requiring early diagnosis, physiological stabilisation and surgical correction. Jejunoileal atresia presents particular challenges because of variations in anatomical type, proximal bowel dilatation, distal bowel calibre, intestinal viability and residual bowel length. Surgical treatment aims to establish functional intestinal continuity while preserving adequate bowel length and minimising postoperative complications. Despite improvements in neonatal intensive care and nutritional support, postoperative obstruction, feeding difficulties, infection and repeat surgery remain important concerns. Yeung et al. (2016) [1], in a series of 43 neonates undergoing primary repair, reported that 23.3% required reoperation, highlighting the continuing burden of mechanical and functional intestinal complications. Primary anastomosis restores intestinal continuity during the initial operation and avoids the need for subsequent stoma closure. However, its suitability depends on the infant’s physiological condition and intraoperative findings. Staged management provides temporary intestinal diversion followed by delayed restoration of continuity, but exposes infants to stoma-related morbidity and additional surgery. Peng et al. (2019) [2] compared three procedures in 105 neonates with severe jejunoileal atresia. They reported fewer complications and reoperations following the Bishop-Koop procedure than following primary anastomosis, demonstrating that outcomes in anatomically complex disease may differ from those in uncomplicated atresia. Assessment of staged management requires consideration of complications arising after both the initial operation and subsequent stoma closure. Eeftinck Schattenkerk et al. (2021) [3] studied 80 patients and found more severe postoperative complications following the combined enterostomy procedures than following primary anastomosis. Their findings emphasised the importance of evaluating the complete treatment pathway. Fung et al. (2023) [4], in a retrospective study of 62 neonates, found that primary anastomosis was associated with shorter hospitalisation, although several other postoperative outcomes were comparable between the groups. These findings also illustrated how anatomy and clinical condition influenced operative selection. More recently, Bethell et al. (2025) [5] reported outcomes of 159 infants treated across seven United Kingdom centres. Primary anastomosis was associated with earlier achievement of full enteral feeding, shorter hospital stay and fewer general anaesthetics than enterostomy formation. Nevertheless, observational comparisons require cautious interpretation because infants undergoing staged management may have greater baseline illness or more complex intestinal pathology. AIM To compare postoperative morbidity and clinical outcomes following primary anastomosis and staged surgical management in neonates with jejunoileal atresia. OBJECTIVES 1. To compare the incidence and severity of postoperative complications within 30 days of the initial operation between the two surgical groups. 2. To compare feeding recovery, duration of parenteral nutrition, ventilatory support, cumulative hospital stays, unplanned reoperation and mortality. 3. To assess complications across the complete surgical pathway and identify neonatal and anatomical factors associated with adverse outcomes during six-month follow-up

Source of data Data were obtained from neonates admitted with suspected intestinal atresia to the Department of Paediatric Surgery and the neonatal intensive care unit. Information was collected prospectively from parental interviews, antenatal and referral records, clinical examinations, laboratory and radiological investigations, operative records, anaesthesia charts, intensive care records and follow-up assessments. A structured case-record form was used to document baseline characteristics, operative findings, treatment details and outcomes. Each participant was assigned a unique study identification number. Study design The study was designed as a prospective, hospital-based, comparative observational cohort study involving 120 neonates. Participants were classified according to the surgical strategy undertaken during the initial operation: • Group A-Primary anastomosis: Resection of the atretic segment, where required, followed by restoration of intestinal continuity without a planned stoma. • Group B-Staged surgical management: Initial enterostomy formation, with subsequent planned stoma closure and restoration of intestinal continuity. Treatment selection was made by the attending paediatric surgeon according to clinical condition and operative findings. Random allocation was not performed. Reasons for choosing each strategy were documented to assess confounding by indication. Infants initially treated by primary anastomosis who subsequently required a rescue stoma remained in Group A for the principal analysis, and the rescue procedure was recorded as an adverse outcome. Study location The study was conducted in the Department of Paediatric Surgery, neonatal intensive care unit, operating theatres and paediatric surgical follow-up clinic. Study duration Recruitment was conducted for 12 months. Each infant was followed for six months after the initial operation. The study period therefore included an additional six months after recruitment of the final participant. The recruitment interval required to enrol 120 neonates was determined from the institution’s case volume. Sample size The target sample size was 120 neonates. A planning calculation was based on comparison of two independent proportions Sampling technique Consecutive eligible neonates were enrolled during the recruitment period until the target sample size was reached. A screening register documented eligibility, enrolment and reasons for exclusion. Patients were not selectively recruited to force equal group sizes. Inclusion criteria 1. Neonates aged 0-28 completed days at the initial operation. 2. Jejunal or ileal atresia confirmed during surgery. 3. Initial surgical management by primary anastomosis or enterostomy with planned delayed restoration of continuity. 4. Written informed consent from a parent or legal guardian. Prematurity, low birth weight, multiple atresias, meconium peritonitis and nonlethal associated anomalies were recorded rather than automatically excluded. Exclusion criteria 1. Isolated duodenal or colonic atresia, because their operative management and clinical course differed from those of jejunoileal atresia. 2. Intestinal obstruction caused solely by stenosis, malrotation, Hirschsprung disease, meconium ileus or necrotising enterocolitis without congenital jejunoileal atresia. 3. Previous definitive surgery for atresia at another institution when baseline and operative details could not be reliably obtained. 4. A documented decision for comfort-focused care without corrective surgery. 5. Absence of parental or guardian consent. Bishop-Koop and Santulli procedures, which combined anastomosis with a decompressive stoma, were treated as a separate operative category and were not pooled with conventional staged diversion in the principal two-group comparison. Procedure and methodology Preoperative assessment Following admission, demographic and perinatal information was recorded, including gestational age, sex, birth weight, age at presentation, antenatal findings, delivery details and referral delay. Presenting symptoms, including bilious vomiting, abdominal distension and failure to pass meconium, were documented. Clinical assessment included hydration, temperature, perfusion, respiratory status and evidence of infection. Prematurity, associated anomalies, preoperative ventilatory support and vasoactive support were recorded. Investigations performed as part of clinical care included complete blood count, glucose, serum electrolytes, renal function tests, blood gas analysis and abdominal radiography. Blood cultures, ultrasonography, contrast studies and echocardiography were obtained when clinically indicated. Gastric decompression, fluid and electrolyte correction, thermal care, antibiotics and other stabilisation measures were provided according to institutional neonatal protocols. Intraoperative assessment The site, number and anatomical type of atresia were documented. Jejunoileal atresia was classified as type I, II, IIIa, IIIb or IV. The operative record included proximal bowel dilatation, proximal-to-distal calibre discrepancy, bowel viability, perforation, meconium peritonitis, associated volvulus, length of bowel resected, residual small-bowel length and preservation of the ileocaecal valve. The surgeon’s reason for choosing primary anastomosis or staged management was recorded explicitly. Primary anastomosis group The atretic segment and any nonviable bowel were resected as clinically required, and intestinal continuity was restored. The configuration of the anastomosis, suture technique, number of anastomoses and use of tapering enteroplasty were documented. Staged management group An enterostomy was constructed when the surgeon considered immediate restoration of continuity unsuitable. The stoma configuration and site were recorded. Stoma closure was undertaken after clinical recovery and assessment of nutritional status, distal bowel suitability and operative fitness. Closure timing was individualised and recorded. Infants who died before closure or still had a stoma at six months remained in the staged-management group. Postoperative care Postoperative management followed institutional protocols for respiratory support, analgesia, antimicrobial therapy, fluid replacement and nutrition. Dates of initiation of enteral feeding, achievement of full enteral nutrition and discontinuation of parenteral nutrition were recorded. For infants with enterostomies, daily stoma output, fluid replacement, electrolyte disturbances, peristomal skin injury, prolapse, retraction and other stoma complications were documented. Outcome assessment The primary outcome was the proportion of infants experiencing at least one major postoperative complication within 30 days of the initial operation, defined as Clavien-Dindo grade III-V. Each infant was counted once for this endpoint according to the highest complication grade; individual complications were also reported. Mortality was presented separately. Secondary outcomes included: • Anastomotic leakage, wound infection, wound dehiscence, intestinal obstruction and postoperative sepsis. • Time from initial surgery to first enteral feed and to full enteral nutrition. • Total days of parenteral nutrition and invasive ventilation. • Duration of the initial hospital admission. • Cumulative inpatient days during six months, including stoma-closure admissions. • Planned and unplanned operations, reported separately. • Thirty-day, in-hospital and six-month mortality. • Readmissions, stoma-related morbidity and nutritional status at follow-up. • Restoration of intestinal continuity by six months. Full enteral nutrition was defined as meeting the prescribed nutritional requirement entirely through enteral intake without parenteral nutrition for at least 48 hours. Anastomotic leakage was recorded when supported by operative findings, imaging or enteric discharge attributable to an anastomotic defect. Culture-confirmed sepsis was distinguished from clinically suspected sepsis. Planned stoma closure was not counted as an unplanned reoperation. Complications occurring within 30 days after closure were additionally reported, together with the proportion of infants undergoing closure during follow-up. Follow-up Follow-up assessments were scheduled at approximately two weeks, one month, three months and six months after the initial operation. Additional visits were arranged for stoma care and closure. Feeding, weight, wound healing, stoma status, bowel symptoms, readmissions and further procedures were recorded. Telephone contact was used to establish survival and interim hospitalisation when an in-person visit was missed. Reasons for incomplete follow-up were documented. Sample processing No additional biological samples were collected exclusively for research. Blood samples obtained during clinical care were processed by the hospital laboratory using standard operating procedures. Blood cultures were collected aseptically when infection was suspected, without delaying necessary treatment. Resected intestinal tissue was submitted for histopathological examination according to institutional practice, using appropriate fixation and routine processing. Histopathology findings were recorded when available. Study forms were checked for completeness, coded and entered into a password-protected database. Data discrepancies were resolved against source records. Statistical methods Data were analysed using IBM SPSS Statistics version 26, with additional validated software where required for specialised analyses. Continuous variables were assessed using graphical methods and the Shapiro-Wilk test. Approximately normally distributed variables were summarised as mean ± standard deviation, while skewed variables were presented as median and interquartile range. Categorical variables were expressed as frequency and percentage. Continuous outcomes were compared using the independent-samples t test or Welch’s t test, as appropriate. Skewed outcomes were compared using the Mann-Whitney U test. Categorical outcomes were compared using Pearson’s chi-square test or Fisher’s exact test. The primary outcome was presented as an absolute risk difference and relative risk with 95% confidence intervals. Multivariable logistic regression was used, where the number of outcome events permitted, to assess the association between surgical strategy and major complications. Adjusted odds ratios with 95% confidence intervals were reported. Potential confounders were prespecified from clinical relevance and included birth weight, gestational age, preoperative physiological instability, anatomical severity and perforation or meconium peritonitis. The number of model parameters was limited according to the available events, and highly correlated variables were not entered together without assessment. Deaths before full enteral feeding were not treated as ordinary noninformative censoring. Feeding recovery was assessed with death as a competing event where feasible, and survivor-only summaries were clearly labelled. Hospital stay was interpreted alongside mortality because early death could shorten admission duration. Missing data were described by variable and treatment group. Sensitivity analyses were undertaken where missingness could materially affect the primary result. All tests were two-sided, and p<0.05 was considered statistically significant. Secondary analyses were interpreted as exploratory. Data collection The structured proforma captured five domains: 1. Perinatal characteristics: Gestational age, sex, birth weight, antenatal diagnosis, associated anomalies and referral history. 2. Preoperative condition: Symptoms, age at presentation, hydration, infection status, laboratory findings and organ support. 3. Operative characteristics: Atresia site and type, bowel viability, calibre discrepancy, residual bowel length, procedure and reason for treatment selection. 4. Postoperative outcomes: Complications, feeding milestones, nutritional support, ventilation, hospital stay, reoperations and mortality. 5. Follow-up outcomes: Growth, feeding status, stoma closure, readmissions

Table 1: Baseline neonatal and anatomical characteristics relevant to the comparison of postoperative outcomes Characteristic Primary anastomosis (n=60) Staged management (n=60) Effect estimate (95% CI) Test statistic p value Age at operation, days, mean (SD) 3.4 (1.8) 3.9 (2.1) MD -0.50 (-1.21 to 0.21) t=-1.400 0.164 Gestational age, weeks, mean (SD) 36.8 (2.2) 36.1 (2.6) MD 0.70 (-0.17 to 1.57) t=1.592 0.114 Birth weight, kg, mean (SD) 2.43 (0.48) 2.29 (0.51) MD 0.14 (-0.04 to 0.32) t=1.548 0.124 Male sex, n (%) 35 (58.3) 33 (55.0) RR 1.06 (0.78-1.45) χ²=0.136 0.713 Prematurity, <37 weeks, n (%) 22 (36.7) 26 (43.3) RR 0.85 (0.54-1.31) χ²=0.556 0.456 Low birth weight, <2.5 kg, n (%) 26 (43.3) 30 (50.0) RR 0.87 (0.59-1.27) χ²=0.536 0.464 Complex atresia, type IIIb or IV, n (%) 16 (26.7) 22 (36.7) RR 0.73 (0.43-1.24) χ²=1.386 0.239 Perforation and/or meconium peritonitis, n (%) 13 (21.7) 17 (28.3) RR 0.76 (0.41-1.43) χ²=0.711 0.399 Table 1 compared baseline characteristics between the primary anastomosis and staged management groups, each comprising 60 neonates. The mean age at operation was 3.4±1.8 days in the primary anastomosis group and 3.9±2.1 days in the staged management group, without a statistically significant difference (t=-1.400; p=0.164). Mean gestational age was 36.8±2.2 versus 36.1±2.6 weeks (p=0.114), while mean birth weight was 2.43±0.48 versus 2.29±0.51 kg (p=0.124). Male neonates constituted 58.3% and 55.0% of the respective groups (p=0.713). Prematurity occurred in 36.7% versus 43.3% (p=0.456), and low birth weight in 43.3% versus 50.0% (p=0.464). Complex atresia of type IIIb or IV was present in 26.7% of the primary anastomosis group and 36.7% of the staged management group (RR=0.73; 95% CI: 0.43-1.24; p=0.239). Perforation and/or meconium peritonitis occurred in 21.7% versus 28.3% (RR=0.76; 95% CI: 0.41-1.43; p=0.399). Table 2: Incidence and severity of postoperative complications within 30 days of the initial operation Outcome Primary anastomosis (n=60), n (%) Staged management (n=60), n (%) RR (95% CI) Test statistic p value Any postoperative complication, including death 20 (33.3) 32 (53.3) 0.63 (0.41-0.96) χ²=4.887 0.027 Highest complication severity per infant No complication 40 (66.7) 28 (46.7) 1.43 (1.03-1.98) χ²=4.887 0.027 Minor complication: Clavien-Dindo grade I-II 8 (13.3) 8 (13.3) 1.00 (0.40-2.49) χ²=0.000 1.000 Major complication: Clavien-Dindo grade III-V 12 (20.0) 24 (40.0) 0.50 (0.28-0.91) χ²=5.714 0.017 Individual postoperative complications Surgical-site infection 6 (10.0) 11 (18.3) 0.55 (0.22-1.38) χ²=1.713 0.191 Culture-confirmed postoperative sepsis 8 (13.3) 15 (25.0) 0.53 (0.24-1.16) χ²=2.636 0.104 Intestinal obstruction requiring intervention 5 (8.3) 8 (13.3) 0.63 (0.22-1.80) χ²=0.776 0.378 Wound dehiscence 2 (3.3) 5 (8.3) 0.40 (0.08-1.98) Fisher’s exact† 0.439 Unplanned reoperation within 30 days 7 (11.7) 13 (21.7) 0.54 (0.23-1.26) χ²=2.160 0.142 Thirty-day mortality: Clavien-Dindo grade V 3 (5.0) 6 (10.0) 0.50 (0.13-1.91) Fisher’s exact† 0.491 Table 2 showed that postoperative complications, including death, occurred in 20 neonates (33.3%) following primary anastomosis and 32 (53.3%) following staged management, representing a statistically significant difference (RR=0.63; 95% CI: 0.41-0.96; χ²=4.887; p=0.027). Major complications of Clavien-Dindo grade III-V occurred in 12 neonates (20.0%) in the primary anastomosis group compared with 24 (40.0%) in the staged group (RR=0.50; 95% CI: 0.28-0.91; χ²=5.714; p=0.017). Minor complications occurred in eight neonates (13.3%) in each group (p=1.000). The overall distribution of complication severity differed significantly between groups (χ²=6.118; df=2; p=0.047). Surgical-site infection occurred in 10.0% versus 18.3% (p=0.191), culture-confirmed postoperative sepsis in 13.3% versus 25.0% (p=0.104), intestinal obstruction requiring intervention in 8.3% versus 13.3% (p=0.378), and wound dehiscence in 3.3% versus 8.3% (p=0.439). Unplanned reoperation within 30 days was required in 11.7% versus 21.7% (RR=0.54; 95% CI: 0.23-1.26; p=0.142). Thirty-day mortality was 5.0% following primary anastomosis and 10.0% following staged management (RR=0.50; 95% CI: 0.13-1.91; p=0.491). Table 3: Feeding recovery, supportive care, hospitalisation, unplanned reoperation and mortality Outcome Primary anastomosis Staged management Effect estimate (95% CI) Test statistic p value Recovery outcomes among six-month survivors n=56 n=53 Time to first enteral feed, days, mean (SD) 5.8 (2.1) 8.4 (3.2) MD -2.60 (-3.64 to -1.56) t=-4.986 <0.001 Time to full enteral nutrition, days, mean (SD) 14.6 (5.2) 23.8 (8.7) MD -9.20 (-11.95 to -6.45) t=-6.655 <0.001 Cumulative parenteral nutrition, days, mean (SD) 12.9 (4.8) 21.7 (8.1) MD -8.80 (-11.35 to -6.25) t=-6.852 <0.001 Cumulative invasive ventilation, days, mean (SD) 2.3 (1.4) 3.6 (2.2) MD -1.30 (-2.01 to -0.59) t=-3.658 <0.001 Outcomes among all enrolled neonates n=60 n=60 Cumulative hospital stay through six months, days, mean (SD) 23.6 (7.8) 39.4 (12.6) MD -15.80 (-19.60 to -12.00) t=-8.259 <0.001 At least one unplanned reoperation by six months, n (%) 14 (23.3) 23 (38.3) RR 0.61 (0.35-1.07) χ²=3.165 0.075 Thirty-day mortality, n (%) 3 (5.0) 6 (10.0) RR 0.50 (0.13-1.91) Fisher’s exact 0.491 Six-month mortality, n (%) 4 (6.7) 7 (11.7) RR 0.57 (0.18-1.85) χ²=0.901 0.343 Table 3 demonstrated shorter feeding-recovery and supportive-care durations following primary anastomosis among six-month survivors, comprising 56 infants in the primary anastomosis group and 53 in the staged group. The mean time to first enteral feeding was 5.8±2.1 versus 8.4±3.2 days (mean difference [MD]=-2.60 days; 95% CI: -3.64 to -1.56; t=-4.986; p<0.001). Full enteral nutrition was achieved at 14.6±5.2 versus 23.8±8.7 days (MD=-9.20 days; 95% CI: -11.95 to -6.45; t=-6.655; p<0.001). Cumulative parenteral nutrition was required for 12.9±4.8 versus 21.7±8.1 days (MD=-8.80 days; 95% CI: -11.35 to -6.25; p<0.001), while invasive ventilation lasted 2.3±1.4 versus 3.6±2.2 days (MD=-1.30 days; 95% CI: -2.01 to -0.59; p<0.001). Among all enrolled neonates, cumulative hospital stay was 23.6±7.8 days following primary anastomosis compared with 39.4±12.6 days following staged management (MD=-15.80 days; 95% CI: -19.60 to -12.00; t=-8.259; p<0.001). Unplanned reoperation by six months occurred in 23.3% versus 38.3% (RR=0.61; 95% CI: 0.35-1.07; p=0.075). Neither thirty-day mortality (5.0% versus 10.0%; p=0.491) nor six-month mortality (6.7% versus 11.7%; RR=0.57; 95% CI: 0.18-1.85; p=0.343) differed significantly. Table 4: Cumulative complications across the surgical pathway and factors associated with major adverse outcomes during six-month follow-up Panel A. Outcomes across all procedures occurring within six months Outcome Primary anastomosis (n=60), n (%) Staged management (n=60), n (%) RR (95% CI) Test statistic p value Any postoperative complication, including death 24 (40.0) 38 (63.3) 0.63 (0.44-0.91) χ²=6.541 0.011 Major adverse outcome: Clavien-Dindo grade III-V 14 (23.3) 28 (46.7) 0.50 (0.29-0.85) χ²=7.179 0.007 At least one unplanned readmission 8 (13.3) 18 (30.0) 0.44 (0.21-0.94) χ²=4.910 0.027 Panel B. Unadjusted associations with major adverse outcomes in the pooled cohort (N=120) Neonatal or anatomical factor Major adverse outcome among infants with factor, n/N (%) Major adverse outcome among infants without factor, n/N (%) Crude OR (95% CI) χ² (df=1) p value Prematurity, <37 weeks 24/48 (50.0) 18/72 (25.0) 3.00 (1.38-6.53) 7.912 0.005 Low birth weight, <2.5 kg 26/56 (46.4) 16/64 (25.0) 2.60 (1.20-5.63) 6.028 0.014 Complex atresia, type IIIb or IV 22/38 (57.9) 20/82 (24.4) 4.26 (1.88-9.66) 12.813 <0.001 Preoperative sepsis 14/24 (58.3) 28/96 (29.2) 3.40 (1.35-8.56) 7.179 0.007 Perforation and/or meconium peritonitis 16/30 (53.3) 26/90 (28.9) 2.81 (1.20-6.58) 5.910 Table 4 showed that cumulative postoperative complications across procedures performed within six months occurred in 24 neonates (40.0%) following primary anastomosis and 38 (63.3%) following staged management (RR=0.63; 95% CI: 0.44-0.91; χ²=6.541; p=0.011). Major adverse outcomes occurred in 14 (23.3%) versus 28 (46.7%) neonates (RR=0.50; 95% CI: 0.29-0.85; χ²=7.179; p=0.007). Unplanned readmission was also less frequent in the primary anastomosis group, occurring in eight infants (13.3%) compared with 18 (30.0%) in the staged group (RR=0.44; 95% CI: 0.21-0.94; χ²=4.910; p=0.027). In the pooled cohort, major adverse outcomes occurred in 50.0% of preterm infants compared with 25.0% of term infants (crude OR=3.00; 95% CI: 1.38-6.53; p=0.005), and in 46.4% of infants with low birth weight compared with 25.0% of those without low birth weight (OR=2.60; 95% CI: 1.20-5.63; p=0.014). Complex atresia was associated with adverse outcomes in 57.9% versus 24.4% of infants (OR=4.26; 95% CI: 1.88-9.66; p<0.001). Corresponding proportions were 58.3% versus 29.2% for preoperative sepsis (OR=3.40; 95% CI: 1.35-8.56; p=0.007) and 53.3% versus 28.9% for perforation and/or meconium peritonitis (OR=2.81; 95% CI: 1.20-6.58; p=0.015).

Table 1. Baseline neonatal and anatomical characteristics The supplied tables included 120 neonates, with 60 undergoing primary anastomosis and 60 undergoing staged surgical management. Mean gestational age was 36.8±2.2 versus 36.1±2.6 weeks (p=0.114), and mean birth weight was 2.43±0.48 versus 2.29±0.51 kg (p=0.124). Prematurity affected 36.7% and 43.3% of the respective groups. These proportions were lower than those reported by Schmedding et al. (2021)[1], who identified prematurity in approximately 64% of 435 patients in a German national cohort. Their study also documented substantial associated morbidity, including cardiac anomalies in 21% and abdominal wall defects in 16%. Differences in eligibility criteria, associated anomalies and referral patterns could explain variation between cohorts. Their findings underline the importance of evaluating neonatal vulnerability alongside operative technique. No statistically significant differences were identified in age at operation, sex, prematurity or low birth weight in the supplied comparison. Similarly, Fung et al. (2023)[2], in a retrospective study of 62 neonates, found no significant differences in several demographic characteristics between primary anastomosis and enterostomy groups. Nevertheless, their analysis demonstrated that anatomical location influenced procedure selection. Hillyer et al. (2019)[3] provided further evidence of selection effects: among 92 infants, higher Apgar scores and uncomplicated atresia were associated with receiving primary anastomosis, with adjusted odds ratios of 4.16 and 3.97, respectively. Thus, apparently similar demographic profiles do not establish that treatment groups had equivalent operative risk. In the supplied tables, complex atresia was more frequent in the staged group than in the primary anastomosis group, although the difference was not statistically significant (36.7% versus 26.7%; p=0.239). Perforation and/or meconium peritonitis also occurred more frequently in the staged group (28.3% versus 21.7%; p=0.399). These numerical imbalances could influence postoperative outcomes. Jarkman and Salö (2019)[4], studying 47 children, found that patients with longer hospitalisation had lower birth weight than those with shorter hospitalisation, although their small sample limited identification of independent predictors. Sholadoye et al. (2018)[5] reported that intestinal necrosis, postoperative complications and severe malnutrition were associated with mortality in a Nigerian series of 38 patients. Together, these studies support adjustment for physiological condition, anatomical complexity and nutritional vulnerability rather than relying on nonsignificant baseline p values as evidence of comparability. Table 2. Incidence and severity of postoperative complications within 30 days The hypothetical comparison showed fewer postoperative complications following primary anastomosis than staged management: 33.3% versus 53.3% (RR=0.63; 95% CI: 0.41-0.96; p=0.027). Major complications occurred in 20.0% versus 40.0% (RR=0.50; 95% CI: 0.28-0.91; p=0.017). This direction was consistent with Eeftinck Schattenkerk et al. (2022)[6], who compared 48 patients receiving primary anastomosis with 32 receiving enterostomy. They reported significantly more severe complications following enterostomy treatment and more surgical-site infection, wound dehiscence, adhesive obstruction and short bowel syndrome. However, their evaluation incorporated the combined enterostomy procedures, whereas Table 2 covered only the first 30 days after the initial operation. Their findings are therefore more directly comparable with the cumulative outcomes in Table 4. Primary anastomosis nevertheless carries a meaningful risk of postoperative obstruction and repeat surgery. Yeung et al. (2016)[7] reported reoperation in 10 of 43 neonates (23.3%), occurring between 18 days and four months after primary repair. Functional obstruction accounted for six reoperations, while other indications included stricture, adhesive obstruction and perforation. In the supplied tables, 30-day reoperation occurred in 11.7% after primary anastomosis and 21.7% after staged management (p=0.142). The lower primary-anastomosis percentage compared with Yeung et al. could partly reflect the shorter observation period. Their study illustrates why early technical success should be distinguished from sustained recovery of intestinal function. Evidence from severe atresia does not uniformly favour primary anastomosis. Peng et al. (2019)[8] studied 105 neonates with severe jejunoileal atresia and reported complication rates of 33.3% after Bishop-Koop, 65.3% after primary anastomosis and 71.4% after double-barrelled ileostomy (p=0.003). Reoperation rates were 4.8%, 38.8% and 14.3%, respectively (p<0.001). Likewise, Okello et al. (2024)[9] reported that greater use of Bishop-Koop procedures alongside early enteral nutrition was associated with a reduction in mortality from 71.4% to 45.5% between treatment periods in a resource-limited setting (p=0.03). These findings qualify any general claim that primary anastomosis is superior in all neonates. Bishop-Koop maintains intestinal continuity with a decompressive stoma and should not be treated as equivalent to conventional staged diversion. Individual complications in Table 2, including surgical-site infection, sepsis, obstruction and wound dehiscence, did not differ significantly between groups. Their confidence intervals remained compatible with clinically important differences, so these results do not demonstrate equivalence. Xiong et al. (2024)[10] identified postoperative necrotising enterocolitis in 12 of 180 patients with jejunoileal atresia. Proximal jejunal atresia was more frequent among affected patients than among those without necrotising enterocolitis (58.3% versus 22.6%; p=0.011). This suggests that specific anatomical features can influence individual complications and should be documented separately from the broad classification of surgical strategy. Table 3. Feeding recovery, nutritional support, hospitalisation and mortality Among six-month survivors in the supplied tables, primary anastomosis was associated with earlier first enteral feeding and earlier achievement of full enteral nutrition. Full enteral nutrition was achieved at 14.6±5.2 versus 23.8±8.7 days, a mean difference of -9.20 days (95% CI: -11.95 to -6.45; p<0.001). Bethell et al. (2025) [11], in a multicentre United Kingdom study of 159 infants, similarly reported shorter time to full enteral feeding after primary anastomosis than after enterostomy formation, at 21 versus 60 days (p=0.001). Hospital stay was also shorter, at 28 versus 78 days (p<0.001). Although the direction agrees with the hypothetical comparison, differences in case severity, feeding definitions and statistical summaries prevent direct comparison of the numerical effect sizes. Cumulative parenteral nutrition in the supplied tables lasted 12.9±4.8 days after primary anastomosis and 21.7±8.1 days after staged management (p<0.001). Hillyer et al. (2019) [3] also reported shorter parenteral nutrition duration after primary rather than secondary anastomosis, at 19 versus 74.5 days. However, feeding outcomes depend on postoperative management as well as surgery. Wang et al. (2022) [12], studying 32 patients with intestinal atresia, found that a higher initial feeding volume and faster advancement were associated with earlier achievement of full enteral nutrition (p=0.001) and a lower incidence of intestinal failure-associated liver disease (p=0.032). These findings support documenting feeding protocols when comparing surgical groups. Studies of bowel tapering further demonstrate the contribution of operative details. Aboalazayem et al. (2022) [13] reported a longer median time to full feeding after tapering enteroplasty than after surgery without tapering, at 28 versus 16 days (p=0.022). Conversely, Kurdi et al. (2023)[14] reported earlier full oral intake following antimesenteric sleeve tapering with end-to-end anastomosis compared with end-to-side anastomosis, at 17 versus 20.2 days (p=0.03), together with shorter hospitalisation. These apparently conflicting findings involved different techniques and patient selection. They indicate that “primary anastomosis” encompasses procedures with different functional consequences, making proximal dilatation, calibre discrepancy and bowel-preserving techniques important explanatory variables. The hypothetical cumulative hospital stay was shorter after primary anastomosis, at 23.6±7.8 versus 39.4±12.6 days (p<0.001), whereas six-month mortality did not differ significantly (6.7% versus 11.7%; RR=0.57; 95% CI: 0.18-1.85; p=0.343). This wide interval does not support a firm survival conclusion. Nutritional recovery also extends beyond discharge: Shibuya et al. (2016)[15] found that residual small-bowel length relative to the expected length for gestational age was associated with subsequent catch-up growth. Therefore, time to full feeding should be complemented by growth and nutritional follow-up. In addition, the survivor-only comparisons in Table 3 cannot represent the entire enrolled cohort without acknowledging survivor bias; death before feeding recovery requires separate consideration. Table 4. Cumulative surgical morbidity and factors associated with adverse outcomes Across six months, the supplied tables showed fewer overall complications after primary anastomosis than staged management (40.0% versus 63.3%; RR=0.63; 95% CI: 0.44-0.91; p=0.011), fewer major adverse outcomes (23.3% versus 46.7%; p=0.007) and fewer unplanned readmissions (13.3% versus 30.0%; p=0.027). Wolf et al. (2018)[16] reported complications after enterostomy formation in 80.3% of infants and after closure in 26.0%. Their cohort included several intestinal disorders, particularly necrotising enterocolitis, so these percentages should not be directly transferred to jejunoileal atresia. Nevertheless, the study demonstrates why stoma formation and closure must both contribute to assessment of treatment burden. Longer follow-up is necessary because some complications emerge after the neonatal admission. Røkkum et al. (2023)[17] reported early abdominal reintervention in 11% and late reintervention in 21% of 70 patients, with bowel obstruction the most common indication for late surgery. Despite this morbidity, most respondents had favourable long-term gastrointestinal quality of life. Koenig et al. (2024)[18] reported subsequent intervention for stricture or small-bowel obstruction in 11% of a mixed cohort of 75 infants with intestinal atresia. These studies support reporting both early and later complications, while recognising that a six-month endpoint may miss later obstruction or delayed stoma closure. In the hypothetical pooled analysis, perforation and/or meconium peritonitis was associated with major adverse outcomes (OR=2.81; 95% CI: 1.20-6.58; p=0.015). However, Chan et al. (2017)[19] found no significant differences in reoperation or mortality between neonates with cystic meconium peritonitis-associated atresia and those with isolated atresia. Primary anastomosis was performed in 94% of their patients with cystic meconium peritonitis. Their findings suggest that antenatal meconium peritonitis does not invariably preclude successful primary repair. The combined category in Table 4 may also conceal clinically important differences between antenatal cystic disease and acute perforation with physiological instability. Preoperative sepsis was associated with major adverse outcomes in the supplied analysis (OR=3.40; 95% CI: 1.35-8.56; p=0.007). Prachuapthunyachart et al. (2021)[20] documented immune abnormalities and recurrent infectious complications among children with intestinal atresia referred to an intestinal rehabilitation programme. This selected population differs substantially from an unselected neonatal surgical cohort, but it highlights the multiple contributors to infection-related morbidity. Similarly, anatomical complexity should not be reduced to atresia type alone. Garg et al. (2020)[21], examining resected specimens from 32 patients, found no direct relationship between interstitial cell of Cajal counts and achievement of full enteral nutrition. Their observations caution against assuming that one anatomical or histological feature fully explains postoperative bowel dysfunction. Prematurity, low birth weight and complex atresia were associated with major adverse outcomes in the hypothetical data, with crude odds ratios of 3.00, 2.60 and 4.26, respectively. Otim et al. (2022) [22], in a prospective Ugandan cohort of 76 neonates with bowel obstruction, identified prematurity and fever at admission among predictors of mortality. Their mixed-diagnosis population and shorter follow-up limit direct comparison, but support the relevance of physiological vulnerability and infection. In Table 4, the reported associations remain unadjusted: prematurity, low birth weight, sepsis and anatomical complexity may coexist and influence treatment selection. Consequently, they should be described as associated factors rather than independent predictors, and the apparent advantage of primary anastomosis should not be interpreted as causal.

In this illustrative comparison of 120 neonates with jejunoileal atresia, primary anastomosis was associated with lower overall and major postoperative morbidity than staged surgical management. Among six-month survivors, primary anastomosis was associated with earlier feeding recovery and shorter durations of parenteral nutrition and invasive ventilation. Cumulative hospital stay and unplanned readmission were also lower. However, differences in unplanned reoperation and mortality were not statistically significant. Prematurity, low birth weight, complex atresia, preoperative sepsis, and perforation and/or meconium peritonitis were associated with major adverse outcomes in unadjusted analyses. These findings suggest a potential advantage of primary anastomosis in appropriately selected neonates, but do not establish its superiority across all clinical circumstances. Surgical selection should account for physiological stability, bowel viability, anatomical complexity and available neonatal support. LIMITATIONS OF THE STUDY 1. Observational treatment allocation: Surgical management was selected according to clinical condition and operative findings. Confounding by indication could therefore explain part of the observed differences. 2. Residual baseline imbalance: The staged group had numerically higher proportions of prematurity, low birth weight and complex atresia. Nonsignificant baseline comparisons did not establish equivalent operative risk. 3. Limited sample size: Although 120 neonates were included, relatively few deaths and individual complications occurred. The study had limited precision for these outcomes, and nonsignificant results could not establish equivalence. 4. Unadjusted risk-factor analysis: Associations between neonatal or anatomical factors and adverse outcomes were based on crude odds ratios. These factors could not be considered independent predictors. 5. Survivor bias: Feeding recovery and supportive-care durations were compared only among six-month survivors. Excluding infants who died could affect the magnitude and direction of these comparisons. 6. Variation in perioperative care: Differences in operative technique, surgeon experience, feeding advancement, nutritional support and timing of stoma closure could influence outcomes. 7. Limited follow-up: Six months might not capture delayed stoma closure, late adhesive obstruction, persistent nutritional dependence, growth impairment or neurodevelopmental outcomes. 8. Restricted generalisability: The proposed single-centre population comprised jejunoileal atresia. Findings could not automatically be extended to duodenal or colonic atresia or centres with different resources. 9. Multiple comparisons: Numerous secondary outcomes were assessed without adjustment for multiple testing, increasing the possibility of chance findings.

1. Schmedding A, Hutter M, Gfroerer S, Rolle U. Jejunoileal atresia: a national cohort study. Front Pediatr. 2021;9:665022. doi:10.3389/fped.2021.665022. 2. Fung ACH, Lee MK, Lui MPK, Lip LY, Chung PHY, Wong KKY. Primary anastomosis is the preferred surgical approach for proximal intestinal atresia: a retrospective 20-year analysis. Pediatr Surg Int. 2023;39(1):99. doi:10.1007/s00383-023-05383-4. 3. Hillyer MM, Baxter KJ, Clifton MS, Gillespie SE, Bryan LN, Travers CD, et al. Primary versus secondary anastomosis in intestinal atresia. J Pediatr Surg. 2019;54(3):417-22. doi:10.1016/j.jpedsurg.2018.05.003. 4. Jarkman C, Salö M. Predictive factors for postoperative outcome in children with jejunoileal atresia. Surg J (N Y). 2019;5(4):e131-6. doi:10.1055/s-0039-1697628. 5. Sholadoye TT, Mshelbwala PM, Ameh EA. Presentation and outcome of treatment of jejunoileal atresia in Nigeria. Afr J Paediatr Surg. 2018;15(2):84-7. doi:10.4103/ajps.AJPS_120_16. 6. Eeftinck Schattenkerk LD, Backes M, de Jonge WJ, van Heurn EL, Derikx JP. Treatment of jejunoileal atresia by primary anastomosis or enterostomy: double the operations, double the risk of complications. J Pediatr Surg. 2022;57(9):49-54. doi:10.1016/j.jpedsurg.2021.07.021. 7. Yeung F, Tam YH, Wong YS, Tsui SY, Wong HY, Pang KK, et al. Early reoperations after primary repair of jejunoileal atresia in newborns. J Neonatal Surg. 2016;5(4):42. doi:10.21699/jns.v5i4.444. 8. Peng YF, Zheng HQ, Zhang H, He QM, Wang Z, Zhong W, et al. Comparison of outcomes following three surgical techniques for patients with severe jejunoileal atresia. Gastroenterol Rep (Oxf). 2019;7(6):444-8. doi:10.1093/gastro/goz026. 9. Okello I, Stephens CQ, Kakembo N, Kisa P, Nimanya S, Yap A, et al. Efforts to improve outcomes among neonates with complex intestinal atresia: a single-center low-income country experience. Pediatr Surg Int. 2024;40(1):70. doi:10.1007/s00383-024-05639-7. 10. Xiong X, Lu W, Xing F, Cai Y, Yang J, Yuan Y, et al. Clinical characteristics of postoperative necrotizing enterocolitis in patients with congenital jejunoileal atresia and its risk factors. World J Pediatr Surg. 2024;7(1):e000622. doi:10.1136/wjps-2023-000622. 11. Bethell GS, Hughes B, Varik RS, Chong C, Hall NJ, Muntean A, et al. Management and outcomes of jejunoileal atresia within the United Kingdom. J Pediatr Surg. 2025;60(7):162334. doi:10.1016/j.jpedsurg.2025.162334. 12. Wang J, Xu H, Wang J, Xiao D. Evaluation of postoperative feeding strategies in children with intestinal atresia: a single-center retrospective study. Front Pediatr. 2022;10:953852. doi:10.3389/fped.2022.953852. 13. Aboalazayem A, Ragab M, Magdy A, Bahaaeldin K, Shalaby A. Outcome of tapering enteroplasty in managing jejunoileal atresia. J Indian Assoc Pediatr Surg. 2022;27(6):666-9. doi:10.4103/jiaps.jiaps_1_22. 14. Kurdi M, Mokhtar A, Elkholy M, El-Wassia H, Bamehriz M, Kurdi A, et al. Antimesenteric sleeve tapering enteroplasty with end-to-end anastomosis versus primary end-to-side anastomosis for the management of jejunal/ileal atresia. Asian J Surg. 2023;46(9):3642-7. doi:10.1016/j.asjsur.2023.01.028. 15. Shibuya S, Koga H, Lane GJ, Yamataka A. Factors conducive to catch-up growth in postoperative jejunoileal atresia patients as prognostic markers of outcome. Eur J Pediatr Surg. 2016;26(1):123-7. doi:10.1055/s-0035-1566103. 16. Wolf L, Gfroerer S, Fiegel H, Rolle U. Complications of newborn enterostomies. World J Clin Cases. 2018;6(16):1101-10. doi:10.12998/wjcc.v6.i16.1101. 17. Røkkum H, Johannessen H, Bjørnland K. Perioperative and long-term outcome in patients treated for jejunoileal atresia. J Pediatr Gastroenterol Nutr. 2023;76(4):434-9. doi:10.1097/MPG.0000000000003709. 18. Koenig SM, Russell RT, Quevedo OG, Chen MK. Intestinal atresias: a ten-year evaluation of outcomes. J Surg Res. 2024;296:130-4. doi:10.1016/j.jss.2023.12.015. 19. Chan KW, Lee KH, Wong HY, Tsui SY, Wong YS, Pang KY, et al. Cystic meconium peritonitis with jejunoileal atresia: is it associated with unfavorable outcome? World J Clin Pediatr. 2017;6(1):40-4. doi:10.5409/wjcp.v6.i1.40. 20. Prachuapthunyachart S, Merani S, Cloonan M, Langnas AN, Quiros-Tejeira RE, Vo HD. Immune function and infectious complications in children with jejunoileal atresia. J Pediatr Surg. 2021;56(3):454-8. doi:10.1016/j.jpedsurg.2020.05.036. 21. Garg V, Puri A, Sakhuja P. Novel insights into the histology of jejunoileal atresia and its therapeutic implications. J Pediatr Surg. 2020;55(12):2630-4. doi:10.1016/j.jpedsurg.2020.05.006. 22. Otim P, Elobu EA, Mbiine R, Kakembo N, Komakech D. The etiological spectrum of bowel obstruction and early postoperative outcome among neonates at a tertiary hospital in Uganda. World J Pediatr Surg. 2022;5(4):e000377. doi:10.1136/wjps-2021-000377.