Digital Ink and Surgical Dreams: Perceptions of Artificial Intelligence-Generated Essays in Residency Applications.

INTRODUCTION: Large language models like Chat Generative Pre-Trained Transformer (ChatGPT) are increasingly used in academic writing. Faculty may consider use of artificial intelligence (AI)-generated responses a form of cheating. We sought to determine whether general surgery residency faculty could detect AI versus human-written responses to a text prompt; hypothesizing that faculty would not be able to reliably differentiate AI versus human-written responses. METHODS: Ten essays were generated using a text prompt, "Tell us in 1-2 paragraphs why you are considering the University of Rochester for General Surgery residency" (Current trainees: n = 5, ChatGPT: n = 5). Ten blinded faculty reviewers rated essays (ten-point Likert scale) on the following criteria: desire to interview, relevance to the general surgery residency, overall impression, and AI- or human-generated; with scores and identification error rates compared between the groups. RESULTS: There were no differences between groups for %total points (ChatGPT 66.0 ± 13.5%, human 70.0 ± 23.0%, P = 0.508) or identification error rates (ChatGPT 40.0 ± 35.0%, human 20.0 ± 30.0%, P = 0.175). Except for one, all essays were identified incorrectly by at least two reviewers. Essays identified as human-generated received higher overall impression scores (area under the curve: 0.82 ± 0.04, P < 0.01). CONCLUSIONS: Whether use of AI tools for academic purposes should constitute academic dishonesty is controversial. We demonstrate that human and AI-generated essays are similar in quality, but there is bias against presumed AI-generated essays. Faculty are not able to reliably differentiate human from AI-generated essays, thus bias may be misdirected. AI-tools are becoming ubiquitous and their use is not easily detected. Faculty must expect these tools to play increasing roles in medical education.

The Challenge of Small Intestine Length Measurement: A Systematic Review of Imaging Techniques.

BACKGROUND: Short bowel syndrome is the most common cause of intestinal failure, with morbidity and mortality linked to remanent small intestine length. There is no current standard for noninvasive bowel length measurement. MATERIALS AND METHODS: The literature was systematically searched for articles describing measurements of small intestine length from radiographic studies. Inclusion required reporting intestinal length as an outcome and use of diagnostic imaging for length assessment compared to a ground truth. Two reviewers independently screened studies for inclusion, extracted data, and assessed study quality. RESULTS: Eleven studies met the inclusion criteria and reported small intestinal length measurement using four imaging modalities: barium follow-through, ultrasound, computed tomography, and magnetic resonance. Five barium follow-through studies reported variable correlations with intraoperative measurements (r = 0.43-0.93); most (3/5) reported underestimation of length. US studies (n = 2) did not correlate with ground truths. Two computed tomography studies reported moderate-to-strong correlations with pathologic (r = 0.76) and intraoperative measurements (r = 0.99). Five studies of magnetic resonance showed moderate-to-strong correlations with intraoperative or postmortem measurements (r = 0.70-0.90). Vascular imaging software was used in two studies, and a segmentation algorithm was used for measurements in one. CONCLUSIONS: Noninvasive measurement of small intestine length is challenging. Three-dimensional imaging modalities reduce the risk of length underestimation, which is common with two-dimensional techniques. However, they also require longer times to perform length measurements. Automated segmentation has been trialed for magnetic resonance enterography, but this method does not translate directly to standard diagnostic imaging. While three-dimensional images are most accurate for length measurement, they are limited in their ability to measure intestinal dysmotility, which is an important functional measure in patients with intestinal failure. Future work should validate automated segmentation and measurement software using standard diagnostic imaging protocols.

Social Determinants of Health are Associated With Postoperative Outcomes in Children With Complicated Appendicitis.

INTRODUCTION: Socioeconomic disadvantage has been associated with increased complicated appendicitis rates. Our purpose was to analyze the complex interactions between social determinants of health (SDOH) and postoperative outcomes in pediatric appendicitis. MATERIALS AND METHODS: Children who underwent appendectomy at our institution (1/2015-12/2020) were retrospectively reviewed. We used home addresses to determine composite measures of neighborhood/area-level socioeconomic advantage (Area Deprivation Index [ADI] and Social Deprivation Index [SDI]), and other area-level indicators. We created a novel, composite outcome score computed as a weighted average of eight outcome measures. Feature selection and exploratory factor analysis were used to create a multivariate model predictive of outcomes. RESULTS: Of 1117 children with appendicitis, 20.59% had complicated (perforated) appendicitis. Factor analysis identified two multivariate latent factors; Factor 1 contained SDI, ADI, and % unemployed in the population, and Factor 2 contained % Hispanic and % foreign-born in the population. Low Factor 2 scores (communities with more Hispanic/foreign-born residents) were associated with increased length of stay, more frequent postoperative percutaneous drainage, and increased postoperative imaging. CONCLUSIONS: Interactions between SDOH and pediatric surgical care go beyond the individual patient and suggest that vulnerable populations are exposed to contextual conditions that may impact outcomes. Specifically, neighborhood-level factors, including the prevalence of Hispanic ethnicity and foreign-born individuals, are associated with outcomes in pediatric patients with complicated appendicitis. Reducing disparities in complicated appendicitis outcomes may involve addressing neighborhood-level SDOH through strategic reallocation of healthcare resources and developing targeted interventions to improve access to pediatric surgical care in underserved communities.

Case report of mesenteric strangulation secondary to longstanding ventriculoperitoneal shunt catheter.

BACKGROUND: Ventriculoperitoneal (VP) shunts are the most common treatment for hydrocephalus in both pediatric and adult patients. Complications resulting from the abdominal portion of shunts include tube disconnection, obstruction of the shunt tip, catheter migration, infection, abdominal pseudocysts, and bowel perforation. However, other less common complications can occur. The authors present a unique case of a patient with a longstanding VP shunt presenting with an acute abdomen secondary to knotting of the peritoneal portion of the catheter tubing. CASE DESCRIPTION: A 13-year-old male with past medical history significant for myelomeningocele, requiring ventriculoperitoneal shunt placement at 18 months of age, presented to an outside hospital with chief complaint of abdominal pain. Cross-sectional imaging revealed spontaneous knot formation within the shunt tubing around the base of the small bowel mesentery. He was then transferred to our facility for general and neurosurgical evaluation. His abdominal exam was notable for diffuse distension in addition to tenderness to palpation with guarding and rebound. Given his tenuous clinical status and peritonitis, he was emergently booked for abdominal exploration. He underwent bowel resection, externalization of his shunt, with later re-anastomosis and shunt internalization. He eventually made a full recovery. DISCUSSION: Given the potential for significant bowel loss with this and other shunt-related complications, this case serves as a reminder that even longstanding VP shunts should be considered in the differential diagnosis of abdominal pain in any patient with a shunt.

Current status of surgical management of gastroesophageal reflux in children.

PURPOSE OF REVIEW: Surgical therapy for gastroesophageal reflux disease (GERD) is controversial with considerable debate ranging from the indications for antireflux surgery to surgical technique. This article will attempt to clarify these issues with the most up-to-date information available on the prevalence, pathophysiology, diagnosis, and surgical treatment of GERD in children. Although laparoscopic Nissen fundoplication (LNF) has become the most popular operation performed for pathologic reflux, its superiority over both open surgery and other types of fundoplication is not well established. RECENT FINDINGS: Large retrospective studies suggest LNF has a lower complication rate than open surgery. However, three prospective randomized controlled trials have been published recently which cast doubt on the superiority of LNF and suggest that LNF may have a higher failure rate compared to open fundoplication. Antireflux surgery has higher morbidity and failure rates in infants and in children with neurologic impairment. SUMMARY: Based on the best available evidence, LNF may be less morbid, but have a higher rate of failure than open surgery. Pediatric surgeons should be mindful of the risks and benefits of both approaches to best counsel their patients. Larger prospective randomized controlled trials are needed to determine the best treatments for pediatric GERD.

Ultrasound in preoperative evaluation for dialysis-access placement.

US vessel mapping for vascular access creation has evolved as standard practice in the preoperative evaluation of an ESRD patient. The best yield from this technology is when it is used to compliment physical examination and viewed in realtime by the operating physician. It is important to evaluate both arterial and venous systems during vessel mapping. While minimum diameters have been associated with an increased chance of maturation of an AV access, the quality of the vessel wall also plays an important role in maturation. US has a distinct advantage of being a noninvasive modality that can evaluate both structural and functional aspects of vessels that play an important role in access maturation.

Machine Learning Improves the Accuracy of Trauma Team Activation Level Assignments in Pediatric Patients.

BACKGROUND: The assignment of trauma team activation levels can be conceptualized as a classification task. Machine learning models can be used to optimize classification predictions. Our purpose was to demonstrate proof-of-concept for a machine learning tool for predicting trauma team activation levels in pediatric patients with traumatic injuries. METHODS: Following IRB approval, we retrospectively collected data from the institutional trauma registry and electronic medical record at our Pediatric Trauma Center for all patients (age <18 y) who triggered a trauma team activation (1/2014-12/2021), including: demographics, mechanisms of injury, comorbidities, pre-hospital interventions, numeric variables, and the six "Need for Trauma Intervention (NFTI)" criteria. Three machine learning models (Logistic Regression, Random Forest, Support Vector Machine) were tested 1000 times in separate trials using the union of the Cribari and NFTI metrics as ground-truth (Injury Severity Score >15 or positive for any of 6 NFTI criteria = full activation). Model performance was quantified and compared to emergency department (ED) staff. RESULTS: ED staff had 75% accuracy, an area under the curve (AUC) of 0.73 ± 0.04, and an F1 score of 0.49. The best performing of all machine learning models, the support vector machine, had 80% accuracy, AUC 0.81 ± 4.1e-5, F1 Score 0.80, with less variance compared to other models and ED staff. CONCLUSIONS: All machine learning models outperformed ED staff in all performance metrics. These results suggest that data-driven methods can optimize trauma team activations in the ED, with potential improvements in both patient safety and hospital resource utilization. TYPE OF STUDY: Economic/Decision Analysis or Modeling Studies. LEVEL OF EVIDENCE: II.

In Search of the Truth: Choice of Ground Truth for Predictive Modeling of Trauma Team Activation in Pediatric Trauma.

BACKGROUND: Assigning trauma team activation (TTA) levels for trauma patients is a classification task that machine learning models can help optimize. However, performance is dependent on the "ground-truth" labels used for training. Our purpose was to investigate 2 ground truths, the Cribari matrix and the Need for Trauma Intervention (NFTI), for labeling training data. STUDY DESIGN: Data were retrospectively collected from the institutional trauma registry and electronic medical record, including all pediatric patients (age <18 years) who triggered a TTA (January 2014 to December 2021). Three ground truths were used to label training data: (1) Cribari (Injury Severity Score >15 = full activation), (2) NFTI (positive for any of 6 criteria = full activation), and (3) the union of Cribari+NFTI (either positive = full activation). RESULTS: Of 1,366 patients triaged by trained staff, 143 (10.47%) were considered undertriaged using Cribari, 210 (15.37%) using NFTI, and 273 (19.99%) using Cribari+NFTI. NFTI and Cribari+NFTI were more sensitive to undertriage in patients with penetrating mechanisms of injury (p = 0.006), specifically stab wounds (p = 0.014), compared with Cribari, but Cribari indicated overtriage in more patients who required prehospital airway management (p < 0.001), CPR (p = 0.017), and who had mean lower Glasgow Coma Scale scores on presentation (p < 0.001). The mortality rate was higher in the Cribari overtriage group (7.14%, n = 9) compared with NFTI and Cribari+NFTI (0.00%, n = 0, p = 0.005). CONCLUSIONS: To prioritize patient safety, Cribari+NFTI appears best for training a machine learning algorithm to predict the TTA level.

Assessing the Impact of Neighborhood and Built Environment on Pediatric Perioperative Care: A Systematic Review of the Literature.

CONTEXT: Neighborhood and built environment encompass one key area of the Social Determinants of Health (SDOH) and is frequently assessed using area-level indices. OBJECTIVE: We sought to systematically review the pediatric surgery literature for use of commonly applied area-level indices and to compare their utility for prediction of outcomes. DATA SOURCES: A literature search was conducted using PubMed, Ovid MEDLINE, Ovid MEDLINE Epub Ahead of Print, PsycInfo, and an artificial intelligence search tool (1/2013-2/2023). STUDY SELECTION: Inclusion required pediatric surgical patients in the US, surgical intervention performed, and use of an area-level metric. DATA EXTRACTION: Extraction domains included study, patient, and procedure characteristics. RESULTS: Area Deprivation Index is the most consistent and commonly accepted index. It is also the most granular, as it uses Census Block Groups. Child Opportunity Index is less granular (Census Tract), but incorporates pediatric-specific predictors of risk. Results with Social Vulnerability Index, Neighborhood Deprivation Index, and Neighborhood Socioeconomic Status were less consistent. LIMITATIONS: All studies were retrospective and quality varied from good to fair. CONCLUSIONS: While each index has strengths and limitations, standardization on ideal metric(s) for the pediatric surgical population will help build the inferential power needed to move from understanding the role of SDOH to building meaningful interventions towards equity in care. TYPE OF STUDY: Systematic Review. LEVEL OF EVIDENCE: Level III.

Long-term functional, psychological, emotional, and social outcomes in pediatric victims of violence.

BACKGROUND: To evaluate the long-term functional, psychological, and emotional outcomes in individuals who survived violence-related injuries as children. METHODS: We retrospectively identified all pediatric patients (age <18y at time of injury) treated for a violent traumatic injury (gun-shot wound, stab, or assault) at our institution (1/2011-12/2020). We then prospectively attempted to contact and survey, via telephone, all patients that had reached adulthood (age ≥18y at time of study) using 7 Patient-Reported Outcomes Measurement Information System (PROMIS) instruments and the Primary Care Post Traumatic Stress Disorder (PTSD) screen. RESULTS: Of the 270 patients identified, we attempted to contact 218, successfully contacted 68, and 24 participated in the study. Of participants, 15 (62.5%) sustained gunshot wounds, 8 (33.3%) were stabbed, and 1 (4.2%) was assaulted with a median time from injury of 6.7(3.4) years. Based on PROMIS metrics, Global Physical Health (55.0 vs. 50.0, p = 0.013) and Emotional Support (55.4 vs. 50.0, p = 0.004) were better in participants compared to reference populations. However, a disproportionate number of participants reported substance use in the past 30 days (45.8 vs 13.0%; p < 0.001), 41.7% screened positive for PTSD, and 62.5% requested resources and/or referral for medical care. CONCLUSIONS: Many individuals who survive violent injuries as children continued to experience negative physical and mental outcomes extending into adulthood that required ongoing medical and psychological support. Further resources are needed to better understand the long-term effects of violent injury and to care for the complex needs of this population.

Value Added from Same-day Discharge after Appendectomy for Children with Simple Appendicitis.

Same-day discharge of children after appendectomy for simple appendicitis is safe and associated with enhanced parent satisfaction. Our general pediatric surgeons aimed to improve the rate of same-day discharge after appendectomy for simple appendicitis. Methods: We implemented a clinical practice guideline in September 2019. A surgeon-of-the-week service model and the urgent operating room started in November 2019 and January 2020, respectively. Data for children with simple appendicitis from our academic medical center were gathered prospectively using National Surgical Quality Improvement Program-Pediatric. Patient outcomes before intervention implementation (n = 278) were compared with patients following implementation (n = 264). Results: The average monthly percentage of patients discharged on the day of surgery increased in the postimplementation group (32% versus 75%). Median postoperative length of stay decreased [16.5 hours (interquartile range, 15.9) versus 4.4 hours (interquartile range, 11.7), P < 0.001], and the proportion of patients discharged directly from the postoperative anesthesia care unit increased (22.8% versus 43.6%; P < 0.001). There were no differences in balancing measures, including the return to the emergency department and readmission. Fewer children were discharged home on oral antibiotics after implementation (6.8% versus 1.5%, P = 0.002), and opioid prescribing at discharge remained low (2.5% versus 1.1%, P = 0.385). Conclusions: Using quality improvement methodology and care standardization, we significantly improved the rate of same-day discharge after appendectomy for simple appendicitis without impacting emergency department visits or readmissions. As a result, our health care system saved 140 hospital days over the first 21 months.

Prehospital factors predict outcomes in pediatric trauma: A principal component analysis.

BACKGROUND: Trauma team activation leveling decisions are complex and based on many variables. Accurate triage decisions improve patient safety and resource utilization. Our purpose was to establish proof-of-concept for using principal component analysis (PCA) to identify multivariate predictors of injury severity and to assess their ability to predict outcomes in pediatric trauma patients. We hypothesized that we could identify significant principal components (PCs) among variables used for decisions regarding trauma team activation and that PC scores would be predictive of outcomes in pediatric trauma. METHODS: We conducted a retrospective review of the trauma registry (January 2014 to December 2020) at our pediatric trauma center, including all pediatric patients (age <18 years) who triggered a trauma team activation. Data included patient demographics, prehospital report, Injury Severity Score, and outcomes. Four significant principal components were identified using PCA. Differences in outcome variables between the highest and lowest quartile for PC score were examined. RESULTS: There were 1,090 pediatric patients included. The four significant PCs accounted for greater than 96% of the overall data variance. The first PC was a composite of prehospital Glasgow Coma Scale and Revised Trauma Score and was predictive of outcomes, including injury severity, length of stay, and mortality. The second PC was characterized primarily by prehospital systolic blood pressure and high PC scores were associated with increased length of stay. The third and fourth PCs were characterized by patient age and by prehospital Revised Trauma Score and systolic blood pressure, respectively. CONCLUSION: We demonstrate that, using information available at the time of trauma team activation, PCA can be used to identify key predictors of patient outcome. While the ultimate goal is to create a machine learning-based predictive tool to support and improve clinical decision making, this study serves as a crucial step toward developing a deep understanding of the features of the model and their behavior with actual clinical data. LEVEL OF EVIDENCE: Diagnostic Test or Criteria; Level III.

Outcomes after Child Physical Abuse and Association with Community-Level Social Determinants of Health.

BACKGROUND: Child physical abuse is a significant cause of pediatric injury and death. Previous studies have described disparities in outcomes for physically abused children according to insurance status. We hypothesized that children treated for physical abuse would be more likely to live in neighborhoods with increased socioeconomic deprivation. STUDY DESIGN: We performed a retrospective review of children who were admitted with suspected physical abuse from 2011 to 2021. Home addresses at the time of admission were used to assign an Area Deprivation Index (ADI) of the neighborhood. Clinicopathologic and outcome variables were compared between children from neighborhoods in the top 10th and bottom 90th national neighborhood ADI percentile. Univariate and multivariate logistic models were constructed. RESULTS: One hundred eighty-four children were included for analysis. Children from the top 10th (more impoverished) ADI percentile presented with more severe injuries, had higher area injury scores in the abdomen and extremities, and required admission to the intensive care unit more often, compared with children from the bottom 90th ADI percentile (all p Values <0.05). Children from high ADI neighborhoods were more likely to be discharged to a different caretaker than children from low ADI neighborhoods (71% caretaker change vs 49% caretaker change, p = 0.005). Univariate and multivariate logistic regression demonstrated statistically significant association between the ADI score and the need for caretaker change at the time of discharge (p = 0.004). CONCLUSIONS: Community-level social determinants of health are closely associated with child physical abuse. Child abuse reduction strategies might consider increased support for families with fewer resources and social support systems.

Reduction of surgical site infections in pediatric patients with complicated appendicitis: Utilization of antibiotic stewardship principles and quality improvement methodology.

BACKGROUND: The rate of surgical site infection (SSI) after appendectomy for complicated appendicitis (CA) was high at our children's hospital. We hypothesized that practice standardization, including obtaining intra-operative cultures of abdominal fluid in patients with CA, would improve outcomes and reduce healthcare utilization after appendectomy. METHODS: A quality improvement team designed and implemented a clinical practice guideline for CA that included obtaining intra-operative culture of purulent fluid, administering piperacillin/tazobactam for at least 72 h post-operatively, and transitioning to oral antibiotics based on intraoperative culture data. We compared outcomes before and after guideline implementation. RESULTS: From July 2018-October 2019, 63 children underwent appendectomy for CA compared to 41 children from January-December 2020. Compliance with our process measures are as follows: Intra-operative culture was obtained in 98% of patients post-implementation; 95% received at least 72 h of piperacillin-tazobactam; and culture results were checked on all patients. Culture results altered the choice of discharge antibiotics in 12 (29%) of patients. All-cause morbidity (SSI, emergency department visit, readmission to hospital, percutaneous drain, unplanned return to operating room) decreased significantly from 35% to 15% (p=0.02). Surgical site infections became less frequent, occurring on average every 27 days pre-implementation and every 60 days after care pathway implementation (p=0.03). CONCLUSIONS: Utilization of a clinical practice guideline was associated with reduced morbidity after appendectomy for CA. Intra-operative fluid culture during appendectomy for CA appears to facilitate the selection of appropriate post-operative antibiotics and, thus, minimize SSIs and overall morbidity.