Journal of pediatric surgery最新文献

Background: Necrotizing enterocolitis (NEC) lacks predictive biomarkers and objective early diagnostic markers. Broadband optical spectroscopy (BOS), a transcutaneous noninvasive tool, has previously demonstrated diagnostic specificity and early predictive power for NEC in a mouse model. Here, in continuation of the translational development, we report a first-in-human observational study in premature infants.

Methods: An apparatus for BOS in infants was assembled comprising a handheld probe with broad-spectrum light source coupled to a laboratory-grade spectrometer (ASD LabSpec 4, Malvern Panalytical), with detection range of 350-2500 nm. Inclusion criteria were premature neonates fewer than 36 weeks of gestation without congenital cardiac conditions or abdominal wall defects. Readings were graphed for descriptive comparison and analyzed via support vector machine supervised computer learning.

Results: 96 infants were enrolled over a 3-year period in two large neonatal intensive care units. Four-quadrant abdominal measurements were obtained in under 2 minutes at the time of routine nursing care. Neither patient harm nor any impediments to clinical treatment were noted. Reliable infrared reflectance signals of intra-abdominal intestine were acquired from infants of all Fitzpatrick skin tones. Ten infants developed Bell Stage 2 (moderate) or 3 (severe) NEC during the study and another four had a spontaneous intestinal perforation or other identified intraabdominal process. BOS measurements taken during active NEC episodes were visibly different than same-infant readings and could be retrospectively identified with over 90% sensitivity and specificity in a machine learning model.

Conclusions: BOS is a safe, feasible, noninvasive technology for point-of-care assessment of NEC. The presence of detectable signal changes in premature infants with Bell Stage 2 or 3 NEC suggests that BOS shows promise as a modality of screening or early detection in this vulnerable population.

Introduction: Choledochal cyst (CC) excision is indicated to avoid developing complications. There is no consensus on postoperative follow-up investigations, duration, or timing. At our institution, patients are followed for 5 years after CC excision with ultrasound and liver laboratory tests. We aimed to evaluate both the post-operative outcomes after pediatric CC excision, focusing on long-term complications, and the adherence to our postoperative follow-up protocol.

Methods: A retrospective chart review of children who underwent a CC excision at our institution between January 1986 and March 2025 was performed. Complications occurring over 5 years after surgery were defined as late complications. Comparisons were made between management eras (early: 1986-2001 vs. modern: 2002-2025). p<0.05 was considered significant.

Results: While 57 children underwent CC excision, 49 (86%) had postoperative follow-up and were included in outcomes analyses. Median follow-up duration for the entire cohort was 2.4 years (34 days-16.2 years). Out of 50 patients who had undergone surgery over 5 years ago, 22 patients were followed for of at least 5 years (overall adherence to our institutional postoperative follow-up protocol: 44%). Median follow-up for these 22 patients was 12.0 years (5.4-16.2). Eight patients (8/22, 36%) suffered from late complications. Additionally, 8 patients were transitioned to an adult provider.

Conclusion: While most patients have good long-term outcomes after pediatric CC excision, major complications can still occur later than five years after surgery. Studies focusing on late outcomes, including after transition to adult care, are needed to determine the optimal long-term surveillance after childhood CC excision.

Level of evidence: III.

Introduction: Hirschsprung disease (HD) is a congenital disorder marked by aganglionosis of intestinal nerve plexuses, leading to bowel obstruction. Determining the extent of the aganglionic segment is essential for successful surgical correction. Consequently, the extent of aganglionosis must be examined histologically. Fluorescence confocal microscopy (FCM) may serve as an alternative diagnostic modality. Using a laser to scan the unfixed specimen, it enables rapid intraoperative imaging without causing tissue alteration. This study assessed the feasibility and diagnostic accuracy of FCM in evaluating Hirschsprung disease in pediatric patients.

Materials and methods: Patients undergoing rectal biopsy, ostomy closure, or transanal endorectal pull-through (TERPT) at our center from January to August 2024 were included. Tissue samples were imaged by FCM followed by frozen section and formalin-fixed paraffin-embedded (FFPE) histology. Image quality, assessability of submucosal and myenteric plexuses, and diagnostic agreement with FFPE sections were investigated.

Results: A total of 34 samples from 8 patients were analyzed by FCM and FFPE, the 24 TERPT-derived samples were additionally analyzed by frozen section. Diagnostic agreement with FFPE was 78.5% for FCM and 86.9% for frozen sections (p=0.49). Among fully assessable samples, FCM achieved 80.9% agreement with FFPE compared to 86.4% for frozen sections.

Conclusion: FCM is an easy-to-implement method for identifying normal and affected bowel in HD. It is used on unfixed specimens, preserving the tissue quality. Its diagnostic accuracy is comparable to that of frozen sections in assessable samples, and it is advantageous in cases where local access to expert pathologists is limited.

Background: Large language models (LLMs) are promising tools in healthcare, particularly for accessing unstructured, text-based electronic health record (EHR) data. This systematic review evaluates the applications of LLMs in the EHR for pediatric and surgical care, model performance compared to traditional methods, and proposed clinical potential to improve healthcare processes, patient outcomes, and overall quality of care.

Methods: A systematic search of ten databases from inception to November 2024 was conducted according to PRISMA guidelines, focusing on LLM use in EHRs. Two reviewers independently screened studies for inclusion, with a third reviewer resolving conflicts. Risk of bias was assessed using PROBAST.

Results: Among 4,326 identified studies, 44 met the inclusion criteria - 30 (68.2%) in all surgical specialties, 3 (6.8%) in pediatric surgical subspecialties, and 11 (25%) in pediatrics. Most studies (59.1%) were published in 2024. LLM types included Bidirectional Encoder Representations from Transformers (BERT) and their variants (52.3%), ChatGPT (29.5%), and other models (18.2%). Most studies (90.9%) relied solely on retrospective unstructured data, and 40.9% focused on classification tasks. LLMs demonstrated performance improvements in 78.1% of studies with a traditional comparator. Clinical documentation assistance (54.5%) and diagnostic and clinical decision support (36.4%) were the most commonly proposed applications for LLMs.

Conclusion: While LLMs offer opportunities for EHR analysis in pediatric and surgical care, most studies remain early-stage, with notable limitations including limited external validation and lack of evaluation in actual clinical workflows. Future research should prioritize rigorous validation and real-world testing to support their safe and effective use in practice.

Background: Bowel management programs (BMPs) have demonstrated effectiveness in treating severe constipation and encopresis. We aimed to evaluate the transition rate from more invasive to less invasive treatments (e.g. enema to laxative therapy) and identify clinical characteristics that increase likelihood of downgrading treatment over time.

Methods: Patients participating in our institution's BMP (2020-2024) were identified. Those without adequate follow-up after BMP induction were excluded. Follow-up was collected at 6-, 12-, 18-, and 24-month clinical visits. 'Downgrade' was defined as a transition from enemas to laxatives or from laxatives to no medications. Success was defined as <2 stool smears/week. A multivariate Cox regression model was used to ascertain the effect size of independent predictors of ability to downgrade.

Results: Ninety-eight patients were included (63% male). Median age was 8.8 years at BMP induction (IQR 5.7-10.9). Functional constipation (74%), Hirschsprung disease (16%), and anorectal malformation (7%) were the most common diagnoses and 33% had a behavioral diagnosis. 62% of patients were started on large volume enemas. At a median follow-up of 12.9 months (IQR 6.2-24.6), downgrade occurred in 48 patients (49%) with a median time to downgrade of 6.2 months (IQR 5.7 - 11.7).

Conclusion: Approximately half of patients transitioned to less invasive regimens over two years, with earlier transitions most common among functional constipation. Findings support the effectiveness of structured BMPs and inform counseling on the timing and likelihood of downgrading.

Level of evidence: IV.

Purpose: To examine the association between pyloric stenosis development and seasonal variation or feeding practices.

Methods: Ethical approval was obtained. A retrospective chart review was conducted for patients treated for pyloric stenosis at our centre in central Canada between January 1st, 2012 and December 31st, 2022. Sex and age-matched controls were identified as patients admitted within two months of age with pneumonia, inguinal hernia, or orthopedic conditions. Cases and controls were also matched by season of hospital admission. Descriptive statistics and regression analyses were conducted using Microsoft Excel and R.

Results: A total of 296 pyloric stenosis cases and 296 controls were reviewed. Using Poisson regression, the effect of season on pyloric stenosis rates was evaluated while adjusting for year. A statistically significant association between season and PS incidence (p = 0.007) was found. Compared to the winter, pyloric stenosis rates were 1.33 times higher in the fall (95 % CI: 0.93-1.92), 1.41 times higher in the spring (95 % CI: 0.99-2.02), and 1.80 times higher in the summer (95 % CI: 1.29-2.54). Additionally, when controlling for age, sex, and season, there was a significant association between feeding practices and pyloric stenosis development (p < 0.001). Formula-fed infants were 3.7 times more likely to be diagnosed with pyloric stenosis compared to breast milk fed infants, while those fed both breast milk and formula had 1.7 times the odds of diagnosis.

Conclusion: Pyloric stenosis incidence shows significant seasonal variation and feeding practice associations, suggesting potential modifiable risk factors in its development.

Level of evidence: III.

Type of study: Retrospective Case-Control Study.