Dr. Kerri McKie and Dr. Haley Etskovitz - Research Fellowship in Review

All right, good morning and welcome to the second edition of our research fellow presentations, Grand Rounds. Once again, we have two excellent presenters today. Our first presenter is Dr. Carrie McKee. She is doing her residency at Brigham and has been working with Dr. Angelle as a primary mentor, doing clinical research with focus on appendicitis as well as other things. And Dr. Haley Etzkiewicz is a resident at University of Connecticut, and she's been working primarily with Dr. Nandavada with a focus on her spread disease. I really look forward to hearing this. It gives our little insight. We work with Carrie and Haley all the time. We love working with both of them, but I have no idea what they do in their day job. So I look forward to hearing today both about Dr. Nandavada and Dr. Angelle's work as well as, importantly, Haley and Carrie's. So first up is Dr. Etzkiewicz. Dr. Haley Etzkiewicz is a resident at Brigham and has been working with Dr. Nandavada. Dr. Etzkiewicz is a resident at Brigham and has been working with Dr. Etzkiewicz. So morning everyone, my name is Haley. I'm the second year research fellow in the Nandavada lab. Over the past two years, I've had the pleasure to work on various projects with multiple attendings in the Department of Surgery. I've also had the opportunity to elaborate with other departments in the clinical and basic sciences throughout the institution. I've studied intestinal failure with the care team. I've studied VEO, IBD with the music clinic, and I've worked on various projects with the CPMC team. Today, I'm going to focus primarily on my Hirschfung research, which has been the main focus of my time here with Dr. Nandavada. I have nothing to disclose. We've taken multiple approaches to studying the pathophysiology and natural history of Hirschfung disease. The project listed above tell the story of our lab's work that I've helped contribute to, and this is primarily what I'll be discussing today. The question that drives our work is, why do 30 to 50% of Hirschfung patients continue to suffer from enterocolitis and obstructive symptoms after an uncomplicated pull-through procedure? Lorina was the first fellow in the Nandavada lab. She was here just before I was, and when she was here, she spent her time studying the intestinal barrier of patients with Hirschfung disease. She specifically focused on the epithelium, which houses tight junction proteins. Tight junction proteins are transmembrane proteins that dictate ion transport and control and intestinal permeability and cell signaling. The poor pathway, shown in green, is primarily responsible for the transepathelial flux of small ions in water, while the leak pathway, shown in blue, it regulates the passage of larger metabolites and plays a role in lymphocyte passage. The unrestricted pathway is aptly named and is not restricted by molecular size. Previous studies have observed a disruption in these tight junction proteins in inflammatory bowel disease and patients with enterocolitis. For those reasons, this brings us to our first research question in the lab, which was, are there differences in tight junction protein expression between children with Hirschfung disease compared to children without Hirschfung disease? To analyze the expression of these proteins, frozen tissue samples were obtained from the pathology department. We analyzed ganglionic and aganglionic bowel of patients with Hirschfung disease and comparable samples were taken from healthy controls. Tissue microarrays were then produced from these samples and slides were created. Immunistic chemistry was used to assess expression of a variety of tight junction proteins that are listed on the right-hand side. These specific proteins were chosen based on their known roles in the poor and leak pathways. To minimize sampling bias and systemic bias, whole-side scanning was manually performed and fluorescence intensity was measured. The graph on the left-hand side of the screen shows the quantitative comparison organized by experimental condition, which is controlled specimen, Hirschfung anglionic colon and Hirschfung aganglionic colon. And immunohistochemistry images are on the right-hand side with the protein of interest in red and the nuclear stained dapy in blue. We found that clotting too had reduced expression in both the ganglionic and aganglionic tissue samples from Hirschfung patients compared to the controls. However, there was no significant difference in expression between the aganglionic and ganglionic segments from the same patient with Hirschfung disease. Similarly, there's reduced expression of clotting 15 in both ganglionic and aganglionic colon of the Hirschfung specimens as compared to controls with no significant difference between the ganglionic and aganglionic segments. A cluedin expression is also decreased in both the ganglionic and aganglionic segments of Hirschfung samples compared to controls specimens. But again, there was no difference seen between both samples from the Hirschfung patients. So in summary, Lorraine found that clotting too, clotting 15 and a cluedin have decreased expression in ganglionic and aganglionic Hirschfung samples compared to controls, but there's no significant difference in expression between ganglionated and aganglionated segments of the same patient. In conclusion, this led us to believe that the intestinal barrier is inherently different in the residual ganglionic segments in patients with Hirschfung disease. As we know that clotting too and clotting 15 are primarily responsible for sodium and water transport. While a cluedin is primarily helps modulate the leak pathway, it's possible that the reduced expression of these proteins leads to these residual symptoms of constipation and obstruction in these patients. However, we do recognize that these results were taken from a single point in time. These samples were taken from patients who were not acutely ill. They did not have any aracholytis when they were undergoing surgery to obtain these samples. So it made it very difficult to assess if this was a cause or effect of the disease process, which led us into our next question. Are there differences in tight junction protein expression during the progression of Hirschfung disease and the development of an aracholytis? And this is primarily when I entered the lab. So to study the temporal expression of these proteins, we used an EDNR being knockout mouse model, which is the most widely accepted model for Hirschfung disease. There's an absence of ganglia in the distal bowel, just like in humans, and it behaves very similarly from a phenotypic standpoint. Mice present with abdominal dissension and constipation. They reliably develop an aracholytis by the third week of life, and they will die by the fourth week of life. Panel A shows seven day old mouse pups. Panel B shows the gross difference between a two week old wild type mouse, which is the brown mouse, and a two week old EDNR being knockout mouse, which is the white mouse. And Panel C shows the laparotomy that had to be created for tissue collection. We collected proximal and distal colon samples in five heterosigote mice, five homosigote mice, and five wild type mice at one, two, and three weeks. Digital droplet PCR was used to evaluate gene expression. DDPCR allows us to use extremely small tissue samples and multiplex analysis so that we can analyze the expression of multiple proteins from the same samples at the same time. Simultaneously tissue microarrays were produced and slides were created in the same fashion as the previous project, so that quantitative immunohistochemistry could be performed to evaluate spatial protein expression. The tight junction proteins on the right were selected for analysis based on Lorina's previous work. These graphs show our DDPCR results and allow us to analyze changes in gene expression over time within groups and between groups. The Y-axis shown here represents copies of RNA per microliter of sample, and the X-axis shows days at which the samples were collected, so the one, two, and three week mice. The blue bars represent the pooled cohort samples of wild type proximal colon. The red bars represent the wild type distal colon. Green bars are the knockout proximal colon and the purple bars represent the knockout distal colon. So as you can see here, no significant differences were found in the gene expression of clotting two or clotting 15 between groups or within groups at 7, 14 or 21 days. Similarly, no significant differences were found in the gene expression of a cluedin or ZO1 between groups or within groups at 7, 14 or 21 days. This graph represents our immunohistochemistry results. Unfortunately, the tissue from the 7-day mice was too friable, and there was too much tissue loss to be able to include in our samples for immunohistochemistry analysis. So we don't have that data, so we were only able to compare 14 and 21 days. As you can see here, clotting two expression increased significantly in the knockout mouse proximal colon from day 14 to day 21. Clodting two also increased significantly in the wild type mouse proximal colon, but the increase is much more drastic in the knockout mouse. There's no change in clotting two expression in the knockout mouse distal colon, which is also true in the wild type distal colon. There were no significant differences found in clotting 15, occluding or ZO1 expression between groups at 14 or 21 days. In summary, clotting two protein expression is significantly increased in the proximal colon of EDNRB knockout mice compared to wild type mice over time. What we can take away from this is that there seems to be increased clotting two expression that's likely developmental versus inherent to Hirschfung disease pathology, since it's also seen in the wild type. It could also represent the development of NR colitis, which would be a result that fits with other models of inflammation such as IBD, in which clotting two is shown to be increased. It's thought to be a compensatory response to the flux of water and pathogen clearance. And there was no significant difference in any of the other proteins that we looked at. We can determine that the expression of clotting two is dynamic and that it's likely a response to fulminant NR colitis, but we do also recognize the limitations of this model. With the tissue loss and tissue friability, we weren't able to look at all the timelines we wanted to. And there was a very unreliable life expectancy of the hearing model. They did not all survive until four weeks. It actually took many months just to get five and each group of our experimental design. So ultimately, this mouse model is of one specific gene mutation, the EDNRB, which is not the most commonly seen mutation in humans. And we also recognize that a lot of these samples, anything we took at week three or week four, was when these mice were very sick. And we also don't take humans to the operating room when they have fulminant NR colitis. So it makes these results a little bit difficult to translate. And for these reasons, the lab decided to pivot away from using the mouse model and is now focused on using human tissue only, which brings us to our next research question. Can we create a prospective clinical database and tissue bio repository to facilitate herch-brung translational research? In collaboration with the snapper and the brolt labs and the organoid core, we have been building a prospective tissue bank. We obtained fresh tissue samples of ganglionic and a ganglionic colon in herch-brung patients in the operating room at the time of pull-through. We're also collecting tissue samples from patients with anal rectal malformation at the time of P-SARP to be used as healthy control tissue. This bio repository is a well-controlled database with a corresponding red cap that contains clinical information that's obtained at the time of biopsy. We currently have 10 herch-brung patients enrolled and 11 control samples. These tissue samples are being used to isolate stem cells and generate colonoids. We're currently using bulk RNA-seq to look at differential gene expression at day seven when the colonoids are predominantly in a stem-like state. And we're also planning in the future to use a small molecule differentiation protocol to push these colonoids into an enteroendocrine state to study their cell profiles. Generating this bio-bank has led to a lot of diverse collaborations throughout the institution. And I'm sure that Rosa will have some very exciting data to share with you next year during her grand rounds. While we were performing bench work to study the pathophysiology of herch-brung disease, we were simultaneously working on multiple outcomes projects that were focused on the clinical outcomes and experiences of patients with various phenotypes of herch-brung. So today I'm primarily gonna be talking about small and personal herch-brung disease and short segment herch-brung disease as I transition into discussing my clinical projects. Small and personal herch-brung disease only represents 1% of herch-brung cases. And it's defined as having transitioned to ganglionic bowel that is greater than 5 centimeters proximal to the ICB. There's a very small body of literature around this topic and we're lucky at children to be able to be a tertiary care center that gets to see a significant amount of these patients and help elucidate factors that contribute to their outcomes. This brought us to our next research question. What is the natural history of small and personal herch-brung disease and how do patients with intestinal failure secondary to this disease compared to patients with intestinal failure from other ideologies? One of these questions led to a collaboration with the care team and this work was recently presented at APSA. Patients with small and personal herch-brung disease often developed intestinal failure which can lead to P and dependence, liver disease, and clavsy. Residual small bowel length is highly predictive of patient's ability to achieve entral autonomy. However, bowel length was not the most predictive factor of entral autonomy achievement in patients with intestinal failure, secondary to small and intestinal herch-brung disease specifically. So the objective of this study was to evaluate at the influence of factors other than bowel length on entral autonomy in this population. We performed a single center retrospective cohort study of patients in an intestinal rehab program, care. We matched 25 patients with intestinal failure from herch-brung disease to 50 patients with intestinal failure of other ideologies. The Aspen definition of intestinal failure was used which dictates a P and requirement of at least 64 out of 70 consecutive days. Patients were matched on small bowel length and duration of rehabilitation, which was based on their age to account for the influence of age on intestinal rehabilitation outcomes and then a comprehensive chart review was performed to look at short and long-term outcomes of these patients. Gestational age was the only statistically significant demographic difference between groups. This finding is consistent with the higher rate of prematurely seen in necrotizing anarcholytis, which was the ideology of intestinal failure for a large proportion of our control group. Median small bowel length of the entire cohort was 63 centimeters and median duration of follow-up was 6.7 years. 96% of patients with herch-brung disease had a stoma at their most recent follow-up compared to 6% of controls and most of these ostinues were jadunal. Patients with herch-brung disease required parental nutrition for over double the duration of controls with 68% of herch-brung patients having continued P and dependence at their most recent follow-up visit. To understand the impact of small bowel length on these outcomes, we subdivided our cohort into patients who had greater than or less than 50% of their small bowel remaining. 88% of each group had less than 50% of their small bowel and 12% of each group had greater than 50% of their small bowel. When comparing patients with less than 50% small bowel length, so everything highlighted in blue, patients with herch-brung disease required parental nutrition for significantly longer than controls. When comparing patients with greater than 50% of their small bowel length, which is everything highlighted in orange, there was a large difference in median PN duration, but it did not reach statistical significance. This table shows percent entrial nutrition between groups at most recent follow-up and is divided into quartiles. Patients with herch-brung disease were less likely to achieve entrial autonomy regardless of their bowel length category. Most patients with herch-brung disease and less than 50% small bowel lengths, again, highlighted in the blue box, received less than 25% of their calories entrally at most recent follow-up visit. Compared to controls who received greater than 75% of their calories entrally. When comparing patients with greater than 50% of their small bowel lengths, everything in the orange box, we see a similar trend, though this did not reach statistical significance. Small intestinal herch-brung disease patients had a significantly higher rate of clav Z, though no difference was seen in the incidence of intestinal failure associated with her disease between groups. We recognize that our results must be considered within the inherent limitations of a retrospective study and the inability to accurately measure certain confounders like dysmotility and chronic inflammatory states between these groups. In conclusion, we found that despite similar bowel lengths and duration of intestinal rehabilitation, patients with small intestinal herch-brung disease are less likely to wean from parental nutrition and prognosis is particularly poor for patients with less than 50% remaining small bowel. Our future work will focus on trying to better understand the confounders like dysmotility in this population. I'm now going to transition to our project on short segment herch-brung disease. Even though it's very important to take advantage of the influx of rare diseases we see here, we also wanted to use our expertise to study short segment herch-brung disease, which is a much more common phenotype. It represents 85% of herch-brung cases and is defined as the transition to ganglionic bowel distilled to the rectus sigmoid. These patients are cared for by general pediatric surgeons all over the country. As we previously discussed, about 35% of post-op pull-through patients experience residual symptoms. It's not infrequent for these patients to require long-term medical management and subsequent operations. Optimal timing of pull-through procedure remains unknown and the effect on these potential outcomes remains unknown. Operative timing is largely dependent on surgeon and center preferences, and there's not currently any literature that's powered to appropriately answer these questions. Proponents of neonatal pull-through will claim that early surgery eliminates the need for prolonged irrigation and potentially development of pre-operative, excuse me, herch-brung associated n-arcollitis. While proponents of infantile pull-through will say that delayed surgery decreases the risk of anesthesia in the neonatal period and reduces surgical complications as pelvic structures are more developed. Which brought us to our final question. Does timing of pull-through affect the rate of post-operative n-arcollitis and resource utilization? Our study aims to evaluate the incidence of pull-through the incidence of post-operative herch-brung associated n-arcollitis in neonates versus infants who undergo primary pull-through procedure. We also aim to evaluate the resource utilization associated with each of these approaches. To study this question, we proposed a multi-center consortium project to the EPSN which was accepted. Using a consortia will allow us to obtain a large cohort of herch-brung patient and ensure that our results are generalizable. As the site lead, we were responsible for communicating with the 15 participating sites in our study to design the clinical research forms, create a database that could accurately capture all of the different practice patterns and oversee data entry. I've been working on this project for the entire two years of my fellowship and I'm very excited to help answer some of these pressing questions in the herch-brung community. We currently have 16 participating sites who have entered 301 individual patients which already makes this the largest cohort of short segment herch-brung patients ever studied. This is the project timeline. The plan is to complete data collection this summer and hopefully have some preliminary analysis completed in the fall, which hopefully Rosa will also be able to share with you at the end of next year. So, in closing, I would like to thank all of my collaborators that made this work possible. I want to thank my lab team, Rosa, Leah, Lorraine, I want to thank all the attendings in the surgery department and the fellows. I've learned a lot from each and every one of you whether it was in clinic in the OR, on call, Wednesday education or lab meeting. I really appreciate how much you all care and it's very apparent how invested you are in our training. And finally, I would like to especially thank Dr. Nandavada for giving me this opportunity to work with her. Her mentorship and friendship is one of the most influential things that I've gained from this experience and I know we'll go far beyond this fellowship. Thank you for your attention. I'm happy to take any questions. I forgot to mention the meeting we do questions for both at the end. That will be the last time and Dr. Karen McKee. Thank you for the opportunity to present our work. We have no disloyalty. Putting these slides together was an awesome opportunity to reflect on all of the work we've done in the lab in the last two years. And it was difficult to decide what to share with you all today. As you can see, not only was this a fellowship in appendicitis and antibiotics, but also was a fellowship deeply steeped in collaborative work. Through these collaborations, not only have we produced clinically meaningful and practice changing results, but I've also had the immense privilege of being introduced to attending statisticians and residents across the country. While we do not have time to dive deep into every project, I will go through two projects in detail and then provide a roadmap before the antimicrobial stewardship work is taking up. First up on our tour of the appendicitis. Prior work from our lab by Dr. SEMA and on the wire and Dr. Daniel Cameron laid the groundwork for the disease severity assessment and complicated appendicitis through their involvement in an equipped pediatric appendectomy pilot project. This first project I will present aims to validate their preliminary work nearly a decade after the pilot began. Appendicitis is the most common and costly abdominal surgical emergency in childhood with approximately 30% presenting with complicated appendicitis. This parity of contamination has been showed to influence both risks of postoperative complications and resource utilization. Establishment of a standardized severity assessment has several important applications, including postoperative prognastication clinical research reporting and hospital level comparative performance assessment. In 2013, a pilot collaborative of 29 hospitals participating in American College of Surgeons and Esquip Pediatric was established to collect appendicitis-focused variables, including interoperative criteria represented of complicated disease. These included actual immunophecalis, diffuse fibromapural and excitate, visible preparation and interoperatonal abscess. Data analysis from the pilot collaborative demonstrated all four criteria to be associated with worse outcomes and its equipped pediatric subsequently adopted the criteria for severity adjustment in its comparative performance reporting in 2016. However, the validity of these criteria has not been explored outside of the pilot collaborative concern has been raised as to the subjective nature of the criteria and whether they have remained discriminatory without the strict oversight of the collaborative. Furthermore, the influence of interoperative criteria on many clinically relevant and resource intensive outcomes have yet to be explored. With these considerations, the goal of the present study was to provide the most comprehensive and generalizable analysis to date exploring the validity of interoperative criteria and predicting outcomes in children with appendicitis. To accomplish this goal, we performed a retrospective multi-center cohort study of patients undergoing appendectomy using the APS-NISQP Pediatric Appendectomy Procedure Targeted in General Participant Use Files. Outcomes including adverse events such as organ space infection and resource utilization outcomes, including postoperative link of the day. The association of the interoperative findings and the outcomes of interest were explored using multivariable regression models, adjusting for patient characteristics. The final study cohort included nearly 83,000 patients from 148 hospitals. 72% of patients had no findings of complicated appendicitis or uncomplicated appendicitis and 27.9% had at least one interoperative finding of complicated disease. When compared to uncomplicated appendicitis or the absence of any interoperative findings, with each additional finding of complicated appendicitis, there was a significant stepwise increase in the odds of postoperative complications, including percutaneous drain placement, organ space infection and re-operation, as well as a significant stepwise increase in the odds of increased resource utilization, including need for pick line placement, for interal nutrition, hospital revisit, postoperative imaging, hospital length of stay, and operative time. In comparing these three different levels of disease severity to uncomplicated appendicitis, we've validated this discriminatory value of these interoperative findings. The given idea of how these data can be applied at the bedside, the predictive risk of outcomes based on the 16 possible combinations of complicated disease is presented in this table. While not to be reviewed in detail here, the estimated probabilities down in this table provide the most granular bedside prognostication tools to set expectations for patients and caregivers, estimate resource utilization based on severity of disease, and has important policy and reimbursement implications. The results of this analysis support the use of NISQIP to find criteria as the gold standard evidence-based framework for assessing disease severity in children with appendicitis, and further validates from a population-based sample NISQIP's risk-adjusted approach for comparing outcomes amongst the 157 hospitals. In addition to providing a framework for severity assessment in comparative performance reporting, other applications of this framework include refinement of policy to more closely match hospital reimbursement with expected clinical course, development of standardized guidelines for clinical research reporting, and providing more accurate prognostic information for patients and their caregivers. In the realm of antimicrobial stewardship and infection prevention, we collaborated with the American College of Surgeons to perform analyses on four procedures previously identified by my predecessors, including Dr. Sema and Alboar and Dr. Taffern-He, as procedures with the highest rates and greatest burdens of antimicrobial prophylaxis misutilization. When we dove into these projects, we were ready to be the best antimicrobial stewards, advocating for de-implementation and of antibiotics across all four of these groups. However, our coalescence-systemic work was, well, a surprise. Multidisciplinary consensus guidelines derived from adult data from the infectious disease society of America and the surgical infection society recommend, again, prophylaxis use in patients undergoing coalescence-systemic for uncomplicated coalesis. However, the majority of children undergoing coalescence-systemic for uncomplicated coalesis do receive prophylaxis according to data from NISQP pediatric. Despite existing consensus guidelines, there are conflicting data on the benefit of surgical antimicrobial prophylaxis for uncomplicated coalesis, even in the adult literature. One relatively large randomized control trial actually demonstrated a decrease in surgical site infection rates in adults receiving prophylaxis. Additionally, there is a lack of rigorous comparative effectiveness data specific to the pediatric population. Previous analyses have been limited to one single center study of children undergoing elective coalesistectomy, which actually did demonstrate a reduction in surgical site infections associated with prophylaxis use. These considerations, the goal of the study was to compare the rate of surgical site infections in children undergoing coalesistectomy for uncomplicated coalesis, who did and did not receive prophylaxis. So, accomplish this goal, we conducted a retrospective multi center cohort analysis of children who underwent coalesisectomy for uncomplicated coalesis over a three year period at 141 NISQP hospital. Holylathis was defined as uncomplicated if it was present without associated acute coalesisitis, and creatitis, or evidence of common viral duct involvement. Exclusion criteria included administration of intravenous antibiotics for suspected infectious complications prior to coalesisectomy, continuation of prophylaxis following incision closure, cases looked as emergent and concomitant diagnoses of hematologic or oncologic disorders. Our primary outcomes were 30 day postoperative surgical site infection and re-admission rates. The association between outcomes and the use of prophylaxis was estimated using logistic regression models weighted by the inverse probability of treatment with a random effect by hospital to control for hospital level clustering. Over 2200 patients undergoing laffers undergoing coalesisectomy were identified for potential propensity matching. Over 90% received prophylaxis with 28% receiving extended spectrum agents compared to coverage provided by step of the loan alone. The distribution of patient and operative characteristics and the two comparison groups before waiting is presented in this slide. In the unweighted cohort, patients who received antibiotics were more likely to have a higher BMI. Less likely to be white. Less likely to have chronic coalesisitis. More likely to undergo an open surgery. More likely to undergo a collangiogram. An undergo a longer operative procedure. Following waiting, no differences remain between groups in any patient or operative characteristics. The results of the propensity weight cohorts are presented here after balancing on all covariates. Prophylaxis was associated with a decrease in surgical site infection rates from 3.7 to 0.9%. We're flexing a 72% reduction in the adjusted odds for this outcome. Additionally, readmission rates were lower in children who received prophylaxis as compared to those who did not. We next performed a subgroup analysis to determine if there was any benefit to the use of extended spectrum antibiotics. We felt this was a potentially important question to address given that nearly 30% of patients received extended spectrum antibiotics. In the secondary subgroup analysis, the primary outcomes were 38 postoperative surgical site infection and readmission rates similar to the primary analysis. As in our primary analysis, outcomes were compared across treatment groups at the intensity score waiting on covariates, including age, BMI, ASA class, gender, and operative characteristics, including case acuity, operative duration, postoperative diagnosis of conical, status, and the use of interoperative collangiogram. The results of the propensity weighted cohorts are presented here after balancing on all covariates, no clinically or statistically relevant differences were found in surgical site infection rates between groups. Similarly, no differences were found between readmission rates between the groups. In conclusion, these results challenge existing consensus guidelines and support the use of prophylaxis and children undergoing cohes' sector for uncomplicated covalent biases. The use of extended spectrum prophylaxis was common, however, not found to be superior to stuff as all in alone and as such, the use of extended spectrum antibiotics should be avoided. Our other three collaborative analyses with the American College of Surgeons used similar approaches to the approach described in cohes' sector. In our ORCIA Pexy Study, over 2700 patients from 141 NISQP pediatric hospitals met criteria for propensity matching. Overall, 57% of patients received prophylaxis. On the hospital level, prophylaxis used range from 0 to 100% with 9% of hospitals using prophylaxis in all cases, and 21% of hospitals never using prophylaxis. The overall rate of surgical site infection was 1.2%, and the hospital's surgical site infection rates range from 0 to 14%. Provincy matching was used to balance groups on both patient and operative characteristics. In the graph on the right, the black dots represent prophylaxis use, the light gray bars represent case volume at each hospital, and the small dark gray bars represent SSIs. As you can see, although prophylaxis use varied across hospitals, there's no obvious trend with decreasing surgical site infections and increased prophylaxis use. In the patient level, propensity matching analysis, prophylaxis administration for patients undergoing labor stopic ORCIA Pexy is not associated with a decrease in surgical site infection, re-operation, or hospital readmission. These results held despite the significant variation in the use of prophylaxis across 141 NISQP pediatric hospitals. In a complementary hospital level analysis, there was no correlation between rates of prophylaxis utilization and surgical site infection rates among hospitals. As you can see in the graph from left to right, or from no prophylaxis use to near-universal use, there is no obvious trend with decreasing SSIs rates. In our Pylora Miami study, just over 6000 patients from 188 hospitals met criteria, overall 14% were performed open, 46.6% of patients received prophylaxis. On the hospital level, prophylaxis used range from 0 to 100% with 4.1% of hospitals using prophylaxis in all cases, and 3.3% of hospitals never using prophylaxis. The overall rate of surgical site infection was 1.2%, and the hospital SSI rates range from 0 to 9.1%. Propensity matching was again used to balance groups on both patient and operative characteristics. And again, it demonstrated on the graph on the right, although prophylaxis used did vary across hospitals, there was no obvious trend with decreasing SSIs and increased prophylaxis use. In the patient level propensity analysis, there was no decrease in surgical site infection, re-operation, or re-admission rates, overall, associated with prophylaxis administration prior to incident. And in the complimentary hospital level analysis, no correlation was found between increasing rates of prophylaxis utilization and reduction in surgical site infection rates. However, in a propensity matched subgroup analysis of patients undergoing an open Pylora Miami, prophylaxis use was associated with a decrease in surgical site infection. While in the laxroscopic Pylora Miami, there was no differences in surgical site infection with prophylaxis use. Our results suggest a role for selective prophylaxis in those patients undergoing an open Pylora Miami. However, the routine use of prophylaxis for infants undergoing laxroscopic Pylora Miami should be questioned. Finally, a 4A into postoperative antibiotic use. In Pectus X-Govottum, over 3,500 patients from 148 hospitals met criteria. Overall, 54.9% patients received postoperative prophylaxis. On the hospital level, postoperative prophylaxis use range from 0 to 100% with 11.1% of hospitals using prophylaxis in all cases, and 17.7% of hospitals never using prophylaxis. The overall rate of SSI was 1.7%, the hospital SSI rates ranging from 0 to 50%. Prepensity matching was used to balance groups on both patient and operative characteristics. And again, as demonstrated by the graph on the right, although prophylaxis use did vary across hospitals, there's no obvious trend with decreasing surgical site infections and increased prophylaxis use. In the patient level prepenity matching ounces, postoperative prophylaxis administration following Pectus repair was not associated with a decrease in surgical site infection, re-operation, or readmission. In the complimentary hospital level analysis, overall, no correlation was found between increasing rates of postoperative prophylaxis utilization and reduction in surgical site infections among hospitals. This 4A into the benefit of postoperative antibiotics in Pectus repair allows me to seamlessly transition into some of our ongoing collaborative efforts with the Eastern Pediatric Surgery Network on the use of postoperative prophylaxis in gastrointestinal surgery. The goal of the present collaborative study is to fold. First, we will develop a consortium database through the aggregation of each EPSN site existing this quick pediatric data. And second, as proof of concept, we will explore the association between postoperative prophylaxis use in clean contaminated procedures, including forgot, small bowel and colorectal procedures. We anticipate additional analyses of the association of prophylaxis use across all clean and clean contaminated procedures in the pediatric surgery, including cholesterol, spine surgery, or the pedeck in your logic surgery. Currently, we have 12 sites who have transferred their next clip data to Boston Children's. After all the data is collected, our aim is to perform an entity machinosis of the benefit of postoperative prophylaxis in all gastrointestinal surgeries in the upcoming weeks. Finally, a barrier to a lot of the work we do in antimicrobial stewardship is implementation of the recommendations that we make. Collaboration with the pediatric surgery quality collaborative has allowed us the unique opportunities survey and ultimately interview surgeons and antimicrobial stewardship professionals at 31 hospital across the collaborative to gain insight as to how different hospitals have improved appropriate use of prophylaxis. The goal of the present study is to identify factors influencing prophylaxis utilization and barriers to reducing postoperative use. In doing so, we aim to identify and share best practices and strategies for sustained implementation from high performing hospitals within the PSQC. To do so, we conducted a quantitative survey which was distributed to all 31 hospitals participating in the collaborative to better understand the hospital performance trends, the collaborative efforts between surgical and antimicrobial stewardship professionals at participating hospitals, and the qualitative improvement projects that may have been implemented. Based on our survey results, we identified high and low performing hospitals and scheduled follow-up interviews with 14 sites. Using the theoretical domain framework to structure our set of questions, we developed a comprehensive interview guide which includes questions on current practices and culture, strategies for reducing postoperative prophylaxis use, barriers and facilitators, and measures for sustaining impact as well as lessons learned from their experiences. The goal was to summarize the findings from these interviews and develop a toolkit to be dispersed amongst participating hospitals to improve postoperative prophylaxis de-implementation. So concludes our tour through a penicidus antimicrobial stewardship. I really appreciated your attention. Before concluding, I want to say thank you to the number of feeble first 12 our collaborators, including a new Shimaturo and Malvin Coleman at the American College of Surgeons, Chankunizaki at the Eastern Pediatric Surgery Network, Terry Fisher and Kevin Lally at the Pediatric Surgery Quality Collaboratives for supporting these efforts. Second, all of the former Rangel Lab fellows, including Catherine Hie, Jen and Graham, Semen and Andoar and Daniel Cameron, for providing the building blocks for these studies and further guidance via text, phone and email at all hours. Third, to the young Graham, without whom none of this would be possible. While my talk focused a lot on collaboration with different groups, I spend nearly every day collaborating with the young, learning from her statistical expertise, forth and finally to Dr. Angelle, for as dedication to my development as an academic surgeon and for showing me grace and compassion when I needed it most. Thank you. Well, well, Karen Haley, thank you. And as we prepare the session, at the prior session, we see how incredibly fortunate we are. The rest of the world looks to us and how we track such incredible people to spend time with us and to make our faculty look good and vice versa. And you each have had incredible contributions. We do have time. So I'm going to open up to questions and comments from mentors. I just want to extend my compliments from the previous ground rounds, the two of you. Carrie and Haley contribute to the department, not only with their research, only part of which you saw today, but in the elegant way they represented as many conferences and in the way they carry themselves. They are collegiality and they're working those added to the ambience of our lab and our department. Thank you very much. Congratulations. And excellent presentations from both not surprisingly and carry outstanding presentation again, not surprisingly. And it should be so proud of what you accomplished in the past couple of years. And it's nice that you acknowledge Catherine and Seaman, the others, but you know the majority of what you accomplished, you accomplish, you carried out from start to finish all by yourself. And these are the really the projects that have started really get people to think differently by the animaux of first time in a long time. So again, you should be really proud about what you accomplished. I also want to thank you for all the time and effort you put into making these collaborations work. As many people know, oftentimes it's thankless. Right. You have to feed and water these things. It takes a long time to get the IRBs done. I think it's taken us well over a year to get this where it is now. And obviously we still have a lot of work to do to finish those up. But it also sets the tables and the foundation for all future projects and in the future dollars are going to benefit from that. Right. And so that's your legacy. You should be very proud about that as well. And of course, it's been great to watch you mature over the last couple of years. And as is always the case, we're at the point now where the drafts are near perfect. And now you're going back to residency and at the start all over again. But it has been awesome watching you evolve. And again, it's been really a pleasure and a joy acting as your mentor in the past couple of years. And I'm really looking forward to seeing your career develop. So again, thanks again for all you've done for us. I'll echo a lot of what Sean said. It's true of you, Hayley, too. It's amazing to see how research files progress over two years. And I remember meeting you with no prior research experience and look at you now. It's incredible how much you've matured, but from a lot of what you already had. I think you presented the work really beautifully and you were a part of a big pivot in the lab moving into a different area, establishing collaborations, which I think are going to be very fruitful for me and will be for you as well. But to Dr. Fisherman's point, you all step into a much longer progression of academics for faculty and our parts for short periods of time, but lay the foundation for things to come and you did a lot of that. I want to specifically comment on Hayley's incredible leadership with the EPSN project. It was my first time working with that group and Sean was a great mentor to me through that. And you really wrangled in a lot of faculty through a controversial study and established really beautifully done clinical research forms that have not caused a lot of trouble. The data cleaning has been really nice. And I think that comes from good study design. And congratulations on that and I'm invested in your future and excited to see where it goes. Actually I looked at both of you have worked collaboratively with establishing large data sets across many many institutions or utilizing Davis available which is establishing large data sets across many, many institutions or utilizing Davis available, which is, um, up until recently, not something that our department has been recognized for. Dr. Angela has been early in leading in the group and now the collaborative, but Boston Children's used to be a place where we did it ourselves. And I think you both demonstrate how incredibly powerful large data sets and collaborations and motivating others to work together are important and especially valuable for our department for institution, um, the leadership of the demonstration, which has been well recognized throughout our societies. Questions from others, or I called my phone. So. Why the gold letter? Uh, so, so, uh, uncomplicated, colonized and open, pile of my army, but not laparoscopic paramedic. Despite the fact that we put the laparoscopes through sometimes not the most pristine belly buttons, um, little babies. Why those? As a longer question. Uh, so I sort of gloss over this, but a lot of the work that was done by both Catherine and SEMA was working to identify which of these, um, groups of surgeries, people work, quote, misutilizing antibiotics based on sort of previous data that had said we shouldn't be using antibiotics for these, but people still are in the pediatric world. And often it's because there's not studies in the pediatric world or at least studies that are powered to answer that question. And so when we started these projects, we really thought we were going to say to all four of the projects that I presented stop using antibiotics. In fact, Dr. Engel and I had been on the close effect of me front saying no, which we're using it for uncomplicated goal of the size is based on the best, uh, guidelines that were available. Um, but as you importantly pointed out, both in close effect for uncomplicated goal of the isis and open pile our madme the data do show that there is some benefit to giving prophylaxis, which does certainly find the face of recommendations that are out there in previous studies. But when we dove deeper in, especially for co-existectomy after we found these results that were surprising to us, truly we thought we were going to say stop using it, stop using it. We've been saying this for years stop using it and we had to sort of bite our tongues and say, oh my gosh, maybe we've been saying the wrong thing in the consensus guidelines from the surgical infection society might be questions. We dove deep into the adult literature and it actually turns out, um, there are a lot of studies, um, although there are a lot of studies supporting those guidelines saying stop using antibiotics. There was a lot of studies that actually had the opposite finding where there was some benefit. Um, and so it was really interesting to find that that existed in the adult literature and was more correlated with the work that we did. Um, and then similarly in for open pylori monotomy, I think we sort of offhand have been discussing, you know, what might be different with an open incision versus the laparoscopic incision, because as you point out, the concern a lot of times is the involvement of the umbilicus or sticking a port through the umbilicus, that's a dirty site that we aggressively clean before we start any surgery. Um, but you know, maybe in an open pylori monotomy, it's an open incision, but it's still pretty small and you're sort of pulling on the skin and you're pulling it around and maybe damaging that skin and making it, you know, more susceptible to infection is one of the possibilities, but that's really just a pre-supposition on my end. Um, there is studies in the co-assistedectomy world looking at, um, like the bile associated with gallbladder is like what is growing out of the bile and does that correlate with surgical site infections? That doesn't seem to be borne out. So I, not try to answer why, but certainly the data, um, as I said, is the data, especially in co-assistedectomy, it's actually a lot more suggesting that there is a benefit to prophylaxis and these uncomplicated colitis as co-assistedectomies. So, um, as I said, it was surprising to us based on how we sort of went into the study, um, but then in digging deeper it actually makes, makes a lot of sense. So, um, I say this phrase more commonly over the years. Um, in the days of the dinosaurs, um, before there was laparoscopic, um, anything, we used to say for polar anatomy, don't, uh, but, uh, preferably in the opium, don't be feeling for the olive, which is pretty much a lost art, which now, for every case, I always have you guys feel the olive because nobody's ever felt one before in an, in a weight patient. And we used to say don't do that because the, the staff on our fingers, remember, um, is what's going to cause the infection. So, so don't, don't be doing lots of touching right, right before the case. Um, I think that's hogwash. I don't, I don't know if you have data on the actual, I mean, you don't know the bacteria because you get a little, little, erythema, right? It's like you have culture data. So that was probably, um, mythology. Um, it's fascinating that like the bile sterile right, and, and, and, and it's happened. So, um, it just goes to show the value of having large data sets if you can't say why you could say, but it's real. I'm sorry. I took your microphone. That's all right. Uh, again, thank you very much, uh, Carrie and Haley, lovely presentations and your live meeting presentations were always, uh, excellent and further, uh, your presentations, uh, at our national meetings were also greatly appreciated. Uh, I have just a social thing to say, uh, Catherine has arranged a celebration of our, uh, surgical research fellows, uh, on the, uh, 10th floor, uh, of Hale, uh, in the, uh, common area there, uh, at one 30 today, uh, Haley and Carrie will be getting their sign books. Uh, your, uh, certificates will be coming in the mail because there was a typo. And, uh, all of the certificates, so don't, uh, don't, uh, despair. And I just like to welcome everybody here, uh, to that celebration. And, uh, it's really been a fantastic year for our fellows. Thank you all. Thank you. And I think if I'm not, I don't, I don't want to say false expectations, but it's going to be like 90 degrees to that. I think there's actually, I scream at that. Is that right? Yeah, Ben and Jerry is okay. Uh, so Catherine takes care of everybody. I'm telling you, yeah, I'm telling you just went up. So, um, we do have a couple minutes. I, um, as Dr. Nathana Nose, I have like a, uh, action for, um, uh, total clock, first prongs disease, um, through strange prior experience. You, you highlighted, um, in both prior work and your own work. Um, cause versus effect. And in everything with protein expression, right? Is opposed to the presence or absence of a mutation, right? So, in the next study, you can say, mutation is present, not present, but in protein expression, you always have the question of, uh, of, of why, right? How do you anticipate getting, getting past that, right? You could, you could argue that, um, the presence of diversion or the presence of, uh, of, uh, non-relaxing sphincter or the presence of infection, um, or non-infectious narcoleidus. Um, how do you imagine getting, getting past that question of cause effect using, um, using protein expression? Thank you for that question. Um, as you saw, the models that we used in the past, um, were very tenuous and we weren't able to get past that question. I'm hopeful that moving forward into the organoids and starting from the STEM like state to see where everything starts before we push it into differentiation and then using different, different differentiation media to either push it into a disease state or push it into, just differentiating into enter under, and will help us do that. Um, admittedly that will be Rosa's area of expertise, so I'm not the one who specifically growing out those colonoids, but, um, I'm hopeful that by actually using a human tissue model, we will be able to compare the healthy human tissue to the disease model from not only the egg anglionic, but the ganglionic. And if that doesn't answer that question, I think it will at least open up the next question to get us there. Yeah. Well, let me just start by saying thank you to Carrie and Hailey. Great. It's been a great couple of years. Great to watch you guys develop. Hailey has been working with you on the long summer. Her front project and Carrie on a little upset with you, but we'll talk. We'll keep talking about your future. I just want to say from a sort of education and clinical standpoint, how great has been been working with you two in the three fellows who presented in the last couple of weeks for a couple of weeks ago. And reinforced just how important our research fellows are to not just the work that you do in your research fields and with our faculty to progress their academic work, but also with the support you give us clinically and educationally by your participation or educational programs. So thank you to all of you for that and hopefully it's worthwhile for you as well. I do have a question for you, Carrie. You laid a little Easter egg in the conclusions of your talk saying that, you know, the intraport of finding study should also impact considerations around hospital reimbursement. And especially in this day and age where insurance companies are saying they're going to withhold payment or recoup payment for what they consider preventable or avoidable, intraport of complications. It seems like your work sort of turns the table and says, well, actually it was the patient and the physiology and not us. And I wonder if you could comment on that and talk about how that or future work might impact those conversations. That's a great question. I think especially looking at some of the resource utilization data that comes when you're adding multiple interoperative findings yet definitely to be clear the interoperative findings are what when you get in you stick laparoscopin and you see what findings are there in present that the patient brought to the room. And the resource utilization outcomes are the odds or issues on them are significant and very obvious like they're well beyond one. And I think those really do show that, you know, if your postoperative length of stay is increasing dramatically if you have even one finding let alone all four findings or three findings. Additionally needing drain placement or prolonged TPN or pick placement for either 4 pn or 4 antibiotics. That's a very different hospital course than a kid who comes in and gets their appendix out for uncompricated cold sizes and doesn't even get admitted to the hospital. First is even a kid who has a single finding so you just have a visit, maybe just have a visible hole but you don't have diffuse fibrillant period and so where you're seeing the whole abdomen is covered in that. We know clinically watching these kids, those kids stay at different amounts of time. And based on these results it really does show that it does depend on how not only how many you have but which of those findings you do have and it's going to change your postoperative course. And I think importantly that does say that it's an inherent, I don't want to say problem with the disease but it's inherent consequence of the disease that you came in with and understanding that each of those interoperative findings and combinations those interoperative findings are different diseases. Right. They're on the same disease spectrum but they act very differently. And so I think as you sort of alluded to showing the insurance companies or showing just the hospital that it's the disease that the patient came in with. It's the degree of severity that's going to determine how long this patient is going to need either in hospital care or very close follow up or other procedures that are going to have to happen based on the disease that they presented with. And I think that has very different implications for appropriately treating the disease that you're presented with rather than saying like, oh, this kid stayed too long in the hospital because he had complicated appendicitis but you know one kid's complicated appendicitis very different than the other. So I think definitely gives us a good though black work and ground to stand on that this disease severity is very different and it should be reimbursed differently depending on that level of severity, not just it's complicated or uncomplicated like there are different degrees of it as well. Well, terrific. We are definitely a time. I want to again thank you and all of the research shows and particularly the maturity and humbleness by which you recognize your predecessors in laying the groundwork for you along with your mentors and also your laying the groundwork for your successors demonstrates that research is mostly done in a long, accurate time. And we all have the opportunity to accelerate that work and build on the predecessors is quite mature to recognize that as Dr. Jackson said, how celebration up at 130 after indications conference, hail 10 in the winter garden. Strange name winter garden today, but we hope you can all join us for for managers turns out that congratulations to and thank you for incredible work and contributions.