Justesenkofoed6065

49% CT and 21.79% 3D model). Accurate assessment of a separate AAOCA ostium was highest on 3D models (97.40%). Ostial stenosis was more accurately assessed on 3D models (56.41%). When accuracy was separated by subspecialty, CT and 3D models were more accurately assessed by all participants regardless of training. Conclusions Cardiac imagers and congenital cardiothoracic surgeons most accurately assessed AAOCA presence, type and course on cardiac CT and 3D models. 3D models were superior in representation of ostial characteristics. CT and 3D models are overall more accurately assessed by specialists regardless of training.Isolated chylopericardium after cardiac surgery is extremely rare, but potentially fatal. We present an unusual case of late postoperative chylopericardium causing cardiac tamponade 6 weeks after mitral valve repair, tricuspid annuloplasty and left atrial appendage closure via median sternotomy. Emergent pericardiocentesis was performed. Taurocholic acid Microscopic analysis confirmed the presence of chyle. The patient was successfully managed conservatively with oral dietary manipulation and intravenous octreotide.Background Patient-reported reflux is one of the most common complaints after esophagectomy. This study aimed to determine predictors of patient-reported reflux and if a preserved pylorus would protect from symptomatic reflux. Methods A prospective clinical study recorded patient-reported reflux after esophagectomy from August 2015 to July 2018. Eligible patients were at least 6 months from creation of a traditional posterior mediastinal gastric conduit, completed at least one reflux questionnaire, and had the pylorus treated in either a temporary (>100 IU BotoxTM) or permanent manner (pyloromyotomy or pyloroplasty). Results Of the 110 patients meeting inclusion criteria, median age was 65 and 88/110 (80%) were male. BotoxTM was utilized in 15 (14%) patients, pyloromyotomy in 88 (80%), and pyloroplasty in 7 (6%). A thoracic anastomosis was performed in 78 (71%) patients and cervical in 32 (29%). Esophagectomy was performed for malignancy in 105/110 (95%) and 78/110 (71%) patients were treated with perioperative chemoradiation. Multivariable linear regression analysis revealed patient-reported reflux was significantly worse patients with shorter gastric conduit lengths (p=0.02) and patients who did not receive perioperative chemoradiation (p=0.01). No significant difference was found between patients treated with pyloric drainage versus BotoxTM. Conclusions Absence of perioperative chemoradiation therapy and a shorter gastric conduit were predictors of patient-reported reflux after esophagectomy. Although few patients had BotoxTM, preservation of the pylorus did not appear to affect patient-reported reflux. Further objective studies are needed to confirm these findings.Background The presence of significant atrioventricular valve (AVV) regurgitation results in unfavorable conditions that affect the success of single ventricle (SV) multistage palliation. We report our institution's AVV repair experience. Methods We examined incidence of AVV repair in 603 infants who underwent initial SV palliation surgery from 2002-12. We explored patients' characteristics, anatomic and operative details associated with death, transplantation and AVV reoperation. Results Sixty patients received AVV repair during first-stage (n=10), Glenn (n=27), Fontan (n=23). Median age at AVV repair was 6.9 months (IQR 4.2-24.1). Underlying SV anomaly was HLHS (n=30), heterotaxy (n=15), other (n=15). The AVV was tricuspid (n=34), mitral (n=6), common (n=20). Pre-operatively, all patients had AVV regurgitation ≥ moderate and 7 (12%) had ventricular dysfunction ≥ moderate. Post-repair, AVV regurgitation was none/trivial (n=21, 35%), mild (n=21, 35%), ≥ moderate (n=17, 30%). Competing risks analysis showed that 10-years following AVV repair, 18% of patients had undergone AVV reoperation, 26% had died or undergone transplantation, and 56% were alive without subsequent reoperation. Transplant-free survival was 38%, 65% and 100% for AVV repair at first-stage, Glenn or Fontan (p=0.0011) and was 74%, 83% and 56% for tricuspid, mitral and common AVV repair (p=0.344). Factors associated with transplant-free survival were timing of AVV repair, underlying SV anomaly, and systemic ventricle function. Conclusions AVV repair at first-stage surgery and reduced systemic ventricle function are associated with poor outcomes. In those high-risk patients, different approaches that involve initial palliation mode, timing of AVV repair or listing for transplantation might be warranted.Background Venous-arterial extracorporeal membrane oxygenation (VA-ECMO) undoubtedly saves many lives, but is associated with a high degree of patient morbidity, mortality, and resource utilization. We aimed to develop a machine learning algorithm to augment clinical decision making related to VA-ECMO. Methods Patients supported by VA-ECMO at a single institution from May 2011 to October 2018 were retrospectively reviewed. Laboratory values from only the initial 48 hours of VA-ECMO support were used. Data were split into 70% for training, 15% validation and 15% withheld for testing. Feature importance was estimated and dimensionality reduction techniques were utilized. A deep neural network was trained to predict survival to discharge and the final model was assessed using the independent testing cohort. Model performance was compared to that of the SAVE score using a receiver operator characteristic curve. Results Of the 282 eligible adult VA-ECMO patients, 117 (41%) survived to discharge. A total of 1.96 million laboratory values were extracted from the electronic medical record, from which 270 different summary variables were derived for each patient. The most important variables in predicting the primary outcome included lactate, age, total bilirubin, and creatinine. For the testing cohort, the final model achieved 82% overall accuracy and a greater area under the curve (AUC) than the SAVE score (0.92 vs 0.65, p=0.01) in predicting survival to discharge. Conclusions This proof of concept study demonstrates the potential for machine learning models to augment clinical decision making for VA-ECMO patients. Further development with multi-institutional data is warranted.