Purpose To identify the incidence of methicillin-resistant Staphylococcus aureus (MRSA) in Indigenous compared to non-Indigenous children with skin and soft tissue infections (SSTIs). Methodology Following ethical approval, electronic hospital records were retrospectively reviewed to identify microbiological swabs from patients who underwent incision and drainage of SSTIs at Townsville University Hospital between January 2020 and December 2024. Patients under 16 years with their ethnicity data were included. Results 480 children met the inclusion criteria. 243 (51%) identified as Indigenous and 237 (49%) non-Indigenous. MRSA was isolated in significantly more Indigenous (40%, 98/243) than non-Indigenous (19.4%, 46/237) patients (P < 0.001). Indigenous children were 2.8 times more likely to have MRSA (OR: 2.8, 95% CI: 1.99, 3.96). Staphylococcus aureus was cultured in 38.3% (184 / 480) of samples. Other cultured organisms included Streptococcus species (11.6%), mixed skin flora (5.6%) and mixed enteric and anaerobic Bacteria (4.3%). Other multi-resistant organisms included Extended Spectrum Beta-Lactamase producing E. coli (2/480) and Mycobacterium Abscesses (1/480). SSTIs most commonly affected the limbs (28.3%) followed by head and neck (22%), buttocks (14.5%), torso (9.6%), perianal / perineal (9.6%) and groin (5.6%). Most SSTIs were drained by the paediatric surgeons (67%), followed by Orthopaedics (17.5%), with smaller numbers by maxillofacial, ear-nose and throat, and general surgery. Conclusions We identified higher rates of MRSA in Indigenous children with SSTIs compared to non-Indigenous children, justifying empiric MRSA antimicrobial coverage for Indigenous children. Future prospective studies should include follow-up to assess rates of recurrence, side effects of antibiotics and whether the same trend continues into adulthood.

Background: Robotic paediatric surgery is currently in its infancy in Australia and New Zealand. There are questions around incorporation of robotic surgery into healthcare and how this will impact trainee experience and learning. Methods: A survey was distributed to all paediatric surgery trainees in Australia and New Zealand regarding their current access, experience and attitudes toward robotic surgery training. Results: The survey had a 75% response rate (18/24). The majority of paediatric surgical trainees had worked more than six years in paediatric surgery (10/18, 55%). Most (94%) have either not interacted with a robot, or only at a conference/meeting. One trainee (6%) completed a robotics course, five (28%) had sought them out, and only one (6%) had access to a robotic console in their current training hospital. The majority of trainees (84%) believed Australasia was 5-15 years behind the rest of the world, but believed robotic surgery is likely to be critical to the future of paediatric surgery in Australia and New Zealand. More than half of trainees felt robotics should be incorporated into the structured training curriculum at all levels of training (61%). Nine trainees (53%) thought it would have a positive impact on training; six (35%) trainees being neutral on the topic; and only two felt it may impact negatively (12%). Conclusion: Paediatric surgery trainees in Australia and New Zealand currently have no exposure to robotic surgery. This survey highlights the need for robotics training, in variable formats, to be incorporated into the formal curriculum to ensure better exposure and knowledge prior to its introduction in the public healthcare system, which will likely occur over the next ten years. It is critical existing training requirements are met during this transition phase.

Purpose To explore changes in the epidemiology, presentation and management of children with urolithiasis in Western Australia (WA) over the past 40 years. Methods Retrospective study of all children with urolithiasis managed by Paediatric Urology from 2015-2024 in WA. Comparison made with previously published WA data from 1983-1992. Results 165 cases from 2019-2024 were compared with 93 cases from 1983-1992. In the present cohort there was: older age at presentation (mean 10 vs 4 years), a lower proportion of Indigenous children (7% vs 70%) and more frequent metabolic abnormalities (34% vs <1%). The most common presenting symptoms are now pain (80%), vomiting (42%) and haematuria (26%), and not urinary tract infection as in the previous study (0% vs 57%). There was increasing use of tamsulosin (either as an adjunct or for expulsion therapy) and retrograde intrarenal surgery (RIRS), rather than open surgical management. 83% underwent operative intervention. RIRS was the most common intervention, with laser lithotripsy in 34% and retrieval with basket/grasper in 27%. 83% of children achieved total stone clearance after operative intervention. Stone composition was predominantly calcium oxalate (65%) or calcium phosphate (30%), compared to uric acid/urate (51%) previously. Areas identified for future improved care included: education to minimise delayed diagnosis, awareness of likelihood of missed stones on ultrasound (31% of symptomatic stones) and minimising prolonged stent time to reduce associated complications. Conclusions There were distinct changes in the epidemiology of children with urolithiasis in WA over the past 40 years. Presentation patterns and stone composition now approximate those seen in adults. RIRS has been established as a safe and reliable management option in children. There are still many areas for improvement, including earlier diagnosis, advances in intervention/equipment and increasing utility of medical management for stone expulsion.

Background: The spleen is the most commonly injured organ in children. There is variation in care which guidelines aim to address. Most guidelines focus on management in a paediatric hospital, yet 2/3rd present in other settings, with varying resource levels. To address this, we developed a new guideline to support splenic injury management across settings with varying resource levels. We plan to optimise the guideline through an implementation trial. Purpose: To co-develop the components of an implementation optimisation trial, which aims to ensure the new guideline can be successfully, sustainably implemented, in any hospital context in which a child with suspected splenic injury presents. Methodology: Participatory research methods were used to design and run a series of four x two-hour interdisciplinary on-line workshops with 130 clinicians from around NSW, and a group of consumers with trauma system experience. The first workshop used problem trees to elucidate issues with existing guidelines and care of injured children, the second focused on solution identification in each hospital context, and the third on developing interventions to support guideline uptake and use. A draft NSW-context-specific guideline was also presented for feedback. In the final workshop, outcome measures were developed in relation to appropriate care, and consumer and clinician satisfaction. Results: A guideline decision-making tool was developed, along with a suite of interventions to support its use in any context, including: elements of preparation; champions and mentors; design and accessibility; education and training; communication; and external supports. Outcomes of all workshops were presented as a program logic model, forming the basis of a design for a paediatric splenic injury guideline implementation and optimisation trial. Conclusion: Participatory research methods have resulted in a draft splenic injury guideline applicable in multiple contexts and a robust optimisation implementation trial design for sustainable introduction across the trauma system.

Background: CVADs are essential to support long-term intravenous access. Silicone and polyurethane have distinct mechanical properties that may influence complications. Evidence guiding CVAD selection in children is sparse and largely extrapolated from adult studies. This study aimed to synthesise the available paediatric evidence and identify gaps. Methods: A systematic review (1980-2025) was conducted using PRISMA guidelines on studies describing complications of silicone and polyurethane CVADs in children. Proportional meta-analysis for non-comparative studies was conducted and reported as % (95% CI). I2 was used to assess heterogenicity with >50% considered significant. Results: We identified 1,269 articles from 5 databases; 9 studies met the inclusion criteria (6 retrospective, 3 prospective). Complications assessed: -CVAD occlusion: Silicone 2.9% (1.15-5.3); I2 = 49% – Polyurethane 5.7% (2.5-10.1); I2 = 59% -CVAD–associated bloodstream infection (CLABSI): Silicone 16.5% (7.3-28.4); I2 = 89% - Polyurethane 11% (2.5-24.5); I2 = 91% -Local site complications (e.g., wound infection): Silicone 2.0% (0.6-4.1); I2 = 0% - Polyurethane 1.4% (0.2-8.0); I2 = 78% -CVAD‑associated venous thromboembolism: Silicone no data - Polyurethane 2.1% (0.9-14.4); I2 = 74% -Dislodgement/accidental removal: Silicone 7.1% (1.2-17.3); I2 = 80% - Polyurethane no data -Rupture: Silicone 1.0% (0.04-3.1); I2 = 0% - Polyurethane no data -Mechanical complications at removal: Silicone 5.0% (0.4-14.3); I2 = 89% - Polyurethane 10.0% (5.1-16.1); I2 = 65.4% Conclusions: Evidence directly comparing silicone and polyurethane CVADs in children is extremely limited, preventing meaningful statistical comparison. Proportional meta‑analysis of non‑comparative studies showed low but variable complication rates for both materials, with substantial heterogeneity and important gaps in reporting. Current data do not demonstrate clear superiority of either material. Robust prospective comparative studies are needed to determine whether catheter material influences clinical outcomes in paediatric CVADs use.

Purpose: To review contemporary outcomes of congenital diaphragmatic hernia within the Australian setting. Methodology: A retrospective review was performed of all patients with congenital diaphragmatic hernia managed at the Sydney Children’s Hospital from January 2000 – December 2024 (inclusive). Patients who presented outside of the neonatal period were excluded. The medical records of each patient were reviewed with the primary outcome being mortality, and secondary outcomes including use of ECMO and length of stay. Each five year period was analysed separately (P1 2000-2004, P2 2005-2009, P3 2010-2014, P4 2015-2019, P5 2020-2024). Ethics approval was obtained from the local research governance body. Results: 159 patients with CDH were born during the study period (left-sided n= 133, right-sided n= 26). Rates of prenatal diagnosis increased over time (50-61% during P1-3, compared with 72-79% P4-5), with similar increase in numbers of inborn compared with outborn patients. Mortality remained stable over the study period, ranging from 20-24% between time periods. Median length of stay did however decrease over time (P1 = 21 days, P2 = 26, P3 = 15, P4 = 14, P5 = 13). 136/159 (86%) underwent surgical repair at a median age of 3 days, with a trend towards increasing use of thoracoscopy. ECMO was utilised in 7 cases (4.4%). Four patients were commenced on ECMO pre-operatively, two post-operatively, and one did not have attempt at repair. Mortality in patients who required ECMO was 57% (4/7). Conclusion: We report favourable outcomes of CDH in terms of survival compared with published data worldwide, despite low use of ECMO. Although overall survival remained stable across the 25-year period of this study, we observed an increased rate of prenatal diagnosis, as well as a decrease in length of stay for surviving patients.

Introduction: Pelviureteric junction obstruction (PUJO) is the most common cause of pathological antenatally diagnosed urinary tract dilatation. Approximately 50% of infants with P3 dilatation on ultrasound scans (USS) will require surgical intervention. This can be approached from either a laparoscopic or open approach in children <10kg, according to surgeon preference. This study compares both approaches outcomes. Methods: Retrospective study (2016-2024) of infants <10 kg with PUJO undergoing Anderson-Hynes dismembered pyeloplasty and >1 year follow-up. Group A included patients that underwent open pyeloplasty, and Group B patients that underwent laparoscopic pyeloplasty. Follow-up included ultrasound ± MAG3. Length of stay (LOS), complication rates (defined as Clavien-Dindo >3, including re-do), and success rates (defined as improved dilatation on USS and absence of symptoms) were compared using a multivariable logistic regression (95% CI). Results: 115 patients were included in the study – 68 patients underwent 72 open procedures and 47 patients underwent 48 laparoscopic procedures. All children had P3 dilatation. The median age, weight, renal function, and LOS were similar in both groups. The median operative time was 88 minutes in Group A and 138 minutes in Group B. The duration of follow up ranged from 12-114 months in Group A and 12-70 months in Group B. Complications occurred in 7% of open cases and 2% of laparoscopic cases, with the majority being stent issues. The majority of patients still had a degree of dilatation on USS but showed overall improvement. 4 patients in Group A underwent re-do surgery, with an even split between flank and dorsal lumbotomy approaches Conclusions: Although the average operating time is longer for children under 10kg undergoing laparoscopic pyeloplasty compared to open, the results are favorable with a lower proportion having complications or requiring redo surgery.

Introduction: Intraabdominal lymphatic malformations/lymphangiomas are a relatively rare cause of gut volvulus and subsequent bowel obstruction in the paediatric population. These patients inevitably proceed to theatre for resection of the lesion commonly with bowel resection and anastomosis. A subsequent population of children will have intraabdominal lymphatic malformations identified either antenatally, incidentally, or to investigate symptomatology and have non-operative intervention including sclerotherapy, whereby we can obtain an aspirate of the lesion. In a retrospective case series over the past 4 years since our centre has pursued extended genetic testing for lymphatic malformations, we hypothesise that PIK3CA is the gene responsible for encoding these lesions, which to our knowledge will be a novel publication for the genetic basis of intraabdominal lymphatic malformations. This data will hopefully inform genetic targets for alternate (and non-invasive) treatment methods in the future. Methods: This is a case series of at least 16 patients whose histology or cell-free DNA from aspiration was either sent upfront for genetic testing or retrospectively sent for genetic testing and PIK3CA status. 15 patients returned a positive PIK3CA gene result, with one being negative (which reflects genetic testing results for cutaneous lymphatic malformations, widely accepted to be PIK3CA-associated lesions, due to the nature of cell-free DNA genetic testing allowing for more false negatives). Conclusion: We hypothesise that all intraabdominal lymphatic malformations are PIK3CA positive, in-keeping with the nature of overgrowth. This publication will, to our knowledge, represent the first genetic association in such lesions and this potential therapeutic target will therefore broaden treatment options to include non-invasive/medical options.

Negative Pressure Wound Therapy (NPWT) has been introduced into acute paediatric burn care. This presentation summaries a PhD thesis aiming to investigate NPWT variables and identify how they may be optimised to help facilitate implementation and formulate a set of recommendations regarding NPWT application in acute paediatric burn care. Methodology These data for the Thesis were collected from Queensland Children’s Hospital. Various research styles were used to address the thesis aim including two randomised control trials, an ex vivo experimental model, a retrospective cohort study and a case report. Results A retrospective cohort study identified the best cohort for NPWT application was deep partial thickness burns, reducing scar clinic referral probability from 67.2% (95% CI 58.8 – 75.6) to 44.9% (95% CI 35.6 – 54.3%; p<0.001). A ex vivo model demonstrated that pressure changes were observed to the depth of the subcutis only, increasing dressing layers beneath NPWT or using circumferential NPWT dressings decreases the overall delivered pressure. A case report identified that NPWT applied without Mepitel® as a barrier between the skin and dressing may result in hypertrophic scar formation. A PICO™ randomised control trial (RCT) found that PICO™ should be used with ActicoatFlex™ as Acticoat™ had significantly more complications when combined with a PICO™ dressing (p=<0.001, mean difference 62.5%, 95% CI 29.9% - 95.1%). A hand and foot RCT found that there was no difference in dressing burden between NPWT and standard dressings and that overall clinician opinion regarding NPWT improved over the course of the study. 93% of participants reported that they would use NPWT on the hand and/or foot again. Conclusion This thesis through a collation of papers investigates NPWT variables in acute paediatric burn care. A set of recommendations for NPWT application in paediatric burn care has been developed based on these data.

Purpose Pseudoaneurysm formation is a known complication following blunt liver and spleen injury (BLSI) in children. However, there is ongoing debate about the significance and need for screening and treatment. The purpose of this study is to review the incidence and management of pseudoaneurysms following BLSI in children. Methodology A retrospective review of BLSI at the sole, tertiary paediatric trauma centre in Western Australia was conducted. A systematic literature review of all articles from 2015-2025, reporting on paediatric traumatic splenic or hepatic pseudoaneurysms was performed. Results 232 patients with a mean age of 9.5 years were managed with BLSI. There was one mortality secondary to liver injury. Pseudoaneurysm was identified in 11 patients (5%), with eight diagnosed initially and three on delayed imaging. Eight were managed conservatively, and three with angioembolisation (AE). All cases achieved resolution. There were no cases of delayed haemorrhage. Systematic review revealed 160 cases of traumatic pseudoaneurysm in 7 studies. 25% were diagnosed initially and 60% on delayed imaging (15% unknown timing). 47% resolved without intervention, AE was performed in 52% and splenectomy in 1%. Delayed rupture was reported in 25 patients (1.2% of all BLSI). The majority of cases occurred at least one week after injury. However, most children with ‘delayed rupture’ did not require transfusion and there were no reported deaths. 23/25 underwent AE, two underwent splenectomy, and one resolved without intervention. Conclusions: Pseudoaneurysm formation post BLSI is uncommon in children. Delayed rupture is reported in 1.2% of cases in the literature, although there were no cases of death and the majority did not require transfusion. AE can be a safe and effective method for treatment of pseudoaneurysm and should be considered for symptomatic patients. However, the role for routine screening and prophylactic AE of asymptomatic children is less clear.

Background: Recurrent PUJO after index pyeloplasty is rare and has traditionally been suspected based on symptoms including flank pain and vomiting. However, some children may develop recurrent obstruction and loss of function whilst asymptomatic. The frequency of “silent” recurrent PUJO, and its association with loss of renal function remains poorly defined. Methods: We undertook a 25-year retrospective review of all redo pyeloplasties for recurrent PUJO performed at The Children’s Hospital at Westmead. Cases with concomitant obstructive uropathy or incomplete radiological data were excluded. Symptom status at the time of redo surgery and surveillance USS and MAG3 renography data were analysed. Results: Between 2000 and 2025, twenty children underwent a redo pyeloplasty (85% male, 70% left PUJO). Ten children (50%) were asymptomatic at the time, representing silent recurrent PUJO. The remaining 10 children developed one or more symptoms: significant vomiting (5), recurrent ipsilateral flank pain (4), urosepsis (2), distention with a palpable abdominal mass (2) and new hypertension (1). Four children developed symptoms suggestive of obstruction within 72 hours of stent removal. The median time from pyeloplasty to redo pyeloplasty was 5 months (IQR 3.4 – 11.1), with 80% of redo pyeloplasties occurring within 12 months. Prior to redo pyeloplasty, 88% of children had worsening hydronephrosis (median Δ APD +8.0mm (4.0 – 16.0), and 90% children had a loss of renal function on MAG3 (median Δ DRF -10.5% (-22.5 to -7.5)). Clinically significant loss of renal function (>10%) occurred in 50% of both symptomatic and asymptomatic children. Conclusions: Recurrent PUJO is asymptomatic in 50% of cases and can be associated with loss of renal function. Symptom status alone is not a sufficient discriminator of children at risk of functional decline, and hence, early structured imaging surveillance following pyeloplasty is essential.

Purpose: We have followed a weight (Wt) of 2.5 Kg and gestational age (GA) of 34 weeks for stoma closure in neonates based on a previous study presented to RACS 25 years ago. This study was conducted to reset these thresholds to minimise post- operative complications. Methodology: Retrospective study at Waikato Hospital (2015–2025) on neonates undergoing stoma closure for neonatal bowel pathology was conducted. Colostomies due to Hirschsprung disease and anorectal malformations were excluded. Complications were graded using Clavien–Dindo (CD) classification against weight and gestational age at stoma closure. IBM SPSS program was used for statistical analysis. Institutional approval number was 4551P. Results: Sixty three neonates underwent stoma closure. Mean birth Wt was 1.4 Kg (SD 0.96) and mean GA at birth was 29 weeks (SD 5.4). The mean Wt at stoma closure was 3.1 Kg (SD 1.7) and GA 43 weeks (SD 10). Twenty seven (43%) had no complications (CD grade 0). Twelve had CD 1&2 (19%). Twenty four had CD≥3 (38%). The mean Wt was 3.1 Kg (SD 1.7) and GA was 43.4 weeks (SD 10) at stoma closure. Both Wt and GA at closure had a mild negative correlation with no statistical significance (p=0.1 and 0.9 respectively) on Spearman’s correlation. The complications were less over 3 Kg and after 40 weeks GA on scatter plot. Re-feeding into the distal stoma in some neonates may have mitigated early closure. Conclusion: In this single-centre cohort, GA and weight at stoma closure were not statistically associated with the severity of postoperative complications although there was a mild negative correlation. The weight and height threshold could be elevated further to minimise complications in the absence of competing surgical indications such as fluid losses and poor weight gain. These findings support multi-centre collaboration and standardised data capture to set evidence-based benchmarks

Purpose: Despite advances in surgical technique, children with Hirschsprung disease (HD) continue to experience significant morbidity. This study aimed to review contemporary practice patterns and outcomes in the management of HD at a tertiary paediatric surgical centre, with a focus on identifying areas for quality and practice improvement. Methodology: A retrospective cohort audit was conducted of all children diagnosed with HD and managed between 2019 and 2024. Demographic, clinical, operative, and follow-up data were collected using a standardised proforma. Key measures included age at diagnosis, disease extent, operative approach, levelling strategy, stoma use, Hirschsprung-associated enterocolitis (HAEC), perioperative complications, long-term bowel function, and comorbidities. Descriptive statistics were used. Results: Fifty-one patients were included (male:female ratio 2:1). While most patients were diagnosed neonatally (median 6 days), a wide diagnostic range was observed. Short-segment disease predominated (n=41), though extensive disease was present in a significant minority. Stoma formation was common, with 41 patients requiring a pre–pull-through stoma. Median age at definitive pull-through was 146 days, most commonly via laparoscopic Swenson (n=18) or Soave (n=16). HAEC occurred both preoperatively (13%) and postoperatively (27.5%). Frequent practice-related challenges included high-output stomas, stoma prolapse, sphincter dysfunction, and persistent bowel dysfunction, often necessitating further interventions such as Botox therapy, washouts, laxatives, or additional surgery. Syndromic and developmental comorbidities were common. Conclusion: This audit highlights ongoing morbidity in children with Hirschsprung disease and identifies several modifiable practice areas, including diagnostic timeliness, stoma selection and management, levelling strategies, and structured long-term bowel care. Standardised pathways, enhanced multidisciplinary follow-up, and targeted parental education may reduce complications and improve long-term outcomes.

Background: During laparoscopic appendicectomy, the appendix may be removed either directly through the port or within a specimen retrieval bag. Although bag use is often promoted to minimise wound contamination, its routine application, particularly in uncomplicated appendicitis, remains contentious. Existing evidence is inconsistent, and practice varies considerably across centres. This review aimed to synthesise the literature comparing bag versus non bag extraction and to highlight areas requiring further investigation. Methods: A systematic review and meta-analysis (1980–2025) was conducted in accordance with PRISMA guidelines, including studies reporting postoperative outcomes following laparoscopic appendicectomy using either extraction technique. Both comparative and non-comparative designs were eligible. Primary outcomes were total surgical site infection (SSI), extraction port site infection (EPSI), and intra-abdominal abscess (IAA). Random effects meta-analysis was performed, with results presented as risk ratios and (95% CI). Heterogeneity was assessed using I2, with values >50% considered significant. Results: There were 2055 studies identified of which 13 met the inclusion criteria (1 randomised controlled trial (RCT), 1 quasi-RCT, and 11 Comparative studies). Pooled estimates showed: • SSI RR 0.6 (0.2–1.3), p = 0.2, I2 = 91.8% (86.8–94.9); • EPSI RR 0.7 (0.2–2.1), p = 0.5, I2 = 42.4% (0.0–78.8); • IAA RR 0.7 (0.5–1.0), p = 0.06, I2 = 33.8% (0.0–78.2). Conclusions: Evidence comparing bag and non-bag extraction during laparoscopic appendicectomy remains limited and heterogeneous. While pooled data suggest possible differences in IAA in favour of using a retrieval bag, variability in definitions, operative techniques, and reporting standards precludes definitive conclusions. Robust prospective comparative studies are needed to determine whether routine use of a specimen retrieval bag meaningfully reduces postoperative complications.

Purpose: Amongst children with complex appendicitis (CA), post-operative intra-abdominal abscesses (IAA) occur in approximately 18% of cases (1). Peritoneal lavage (PL) is widely used with the expectation that it reduces complications, despite limited supporting evidence. We hypothesised that suction only (SO) without irrigation would be non-inferior to PL in children with CA. Methodology: We conducted a randomised controlled trial (RCT) including three centres, comparing PL and SO in children undergoing laparoscopic appendicectomy for CA. Recruitment occurred from April 2020 to January 2025, but was interrupted during the COVID-19 pandemic. The sequentially numbered, opaque, sealed envelopes (SNOSE) method was used for randomisation. The PL group underwent irrigation with >2000mls 0.9% saline. In the SO <100mls was used. Outcomes included IAA, WI, small bowel obstruction (SBO) and length of stay (LoS). Standardised site-based data collection was performed. Results: In total 195 participants were recruited, including 111 in the PL group and 84 in the SO group. There were no significant demographic differences. Average LoS was 6.5/7 in the PL group and 5.5/7 in the SO group, p<0.01. In the PL group there were 14 IAA detected, compared to 8 in the SO group (P=0.7). Both groups only had one participant that developed a WI (p=1.0), and 2 participants that developed SBO, p=1.0. Conclusion: Our RCT found that PL in children with CA was associated with a statistically significant, although minimal increase in LoS. However, incidence of complicating IAA, WI and SBO were the same amongst PL and SO groups. This supports the hypothesis of non-inferiority of SO compared to PL. Reference: 1. Peter SDS, Sharp SW, Holcomb GW, Ostlie DJ. An evidence-based definition for perforated appendicitis derived from a prospective randomized trial. J Pediatr Surg. 2008;43(12):2242-5. children. J Pediatr Surg. 2012;47(2):317-21.