BACKGROUND:Textbook outcome, defined as survival to discharge without complications or prolonged hospitalization, has garnered increasing interest as a surgical quality metric. The present study used a national database to evaluate the utility of textbook outcomes for hospital benchmarking in adult cardiac surgery. METHODS:All elective admissions entailing coronary artery bypass grafting and/or valve operations were identified from the 2016 to 2022 Nationwide Readmissions Database. Textbook outcome was defined as survival to discharge without cardiac arrest, stroke, prolonged ventilation, renal failure, sepsis, pulmonary embolism, reoperation, or length of stay >14 days. Royston-Parmar models were used to evaluate associations between textbook outcome and 180-day mortality and nonelective readmission. Hierarchical logistic regression was used to identify patient and hospital factors associated with the textbook outcome. Centers with risk-adjusted textbook outcome rates in the lowest decile were designated low textbook outcome hospitals. RESULTS:Among 963,775 patients, 86.0% achieved a textbook outcome. Prolonged hospitalization (58.6%) was the most common reason for a non-textbook outcome. After risk adjustment, textbook outcome patients demonstrated significantly reduced 180-day mortality (hazard ratio, 0.37; 95% confidence interval, 0.33-0.41) and nonelective readmission (hazard ratio, 0.66; 95% confidence interval, 0.65-0.68). Approximately 8.5% of the variation in textbook outcome was attributable to interhospital differences. Low textbook outcome hospitals had lower annual operative volume (median, 107 vs 137 cases/year; P < .001) and were less often teaching hospitals (73.6% vs 79.9%; P = .011). CONCLUSION:This work demonstrates that the textbook outcome quality metric captures clinically meaningful differences in survival and readmissions following elective cardiac surgery. There was substantial center-level variation in textbook outcome rates, suggesting its utility for hospital benchmarking.

Objectives Staged left ventricular recruitment can promote left ventricular growth through controlled volume loading. The optimal pulmonary blood flow source and the need for the bidirectional Glenn during staged left ventricular recruitment remain unclear. This study investigates staged left ventricular recruitment outcomes in patients who underwent staged left ventricular recruitment with or without the bidirectional Glenn. Methods This is a single-institution retrospective review of children undergoing staged left ventricular recruitment (2014-2024). Patients were stratified by pulmonary blood flow source: nonbidirectional Glenn (shunt/Sano upsize or pulmonary artery band alone) and bidirectional Glenn. Results Eighty-eight patients underwent staged left ventricular recruitment with shunt upsizing (n = 11), pulmonary artery band (n = 9), or bidirectional Glenn/super-Glenn (n = 68). Median age at recruitment was 231 days (interquartile range, 495.25). Fundamental diagnosis was hypoplastic left heart syndrome/complex in 44 patients (50%). Median duration of recruitment was 642 days (576.5); duration was significantly lower in the nonbidirectional Glenn group compared with the bidirectional Glenn group (321 [480] days vs 754 [481] days) (P = .02). Forty-five patients (51.1%) achieved biventricular conversion, and 17 patients (19.3%) achieved 1.5-ventricle circulation. For the nonbidirectional Glenn, 90% reached biventricular circulation. Left ventricle indexed end-diastolic volume significantly increased during recruitment for the nonbidirectional Glenn (31.7 [28.6] mL/m2 to 57.7 [34.5]) (P < .001) and bidirectional Glenn (22.7 [14.3] mL/m2 to 48.5 [29.4]) (P < .001). Superior vena cava pressure significantly increased in the bidirectional Glenn group (9.5 [6] mm Hg to 16.5 [5] mm Hg) (P < .001) but not in the nonbidirectional Glenn group. Lymphatic complications occurred in 16% of patients and venovenous collaterals in 65% of patients in the bidirectional Glenn group. No patients in the nonbidirectional Glenn group experienced these problems. Conclusions Shunt upsizing and pulmonary artery band modification can be effective for ventricular volume loading during staged left ventricular recruitment, enabling biventricular conversion while avoiding potential complications of bidirectional Glenn physiology. These strategies should be considered in patients not anticipated to require prolonged recruitment and in patients deemed poor bidirectional Glenn candidates.

BACKGROUND:Cardiogenic shock (CS) is a leading cause of mortality following acute myocardial infarction (AMI). Some patients may require intra-aortic balloon pump (IABP) or percutaneous ventricular assist device (PVAD) placement; however, there is a paucity of standardised algorithms to guide the deployment of each device. The present study evaluated interhospital variation in the use of IABP and PVAD for AMI CS and identified institutional factors associated with hospital-level device preference. METHODS:All non-elective adult hospitalisations entailing AMI and CS were identified within the 2019 Nationwide Readmissions Database. Patients were grouped into IABP, PVAD and non-mechanical circulatory support cohorts. The primary aim was to quantify the degree of interhospital variation in the use of IABP and PVAD. Escalation to extracorporeal membrane oxygenation (ECMO), left ventricular assist device (LVAD) implantation, length of stay and hospitalisation costs were secondarily assessed. Hospital factors, such as percutaneous coronary intervention (PCI) volume and safety net status, were also analysed. RESULTS:Among 53 903 patients, 23.4% received IABP, and 12.5% received PVAD. After adjustment for patient factors, approximately 13% (11-14%) of variation in IABP use and 18% (15-20%) of PVAD use were attributable to centre-level differences. High-PVAD hospitals had higher annual PCI volume (257 (185-369) vs 204 (148-276) cases/year, p=0.032) and were more commonly safety net institutions (27.4% vs 11.3%, p=0.023), compared to high-IABP hospitals. Patients treated at high-IABP and high-PVAD hospitals faced similar length of stay (β -0.16, 95% CI -1.82 to 1.49) and costs (β -$3500, 95% CI -16 600 to 9600). Those at high-PVAD hospitals had lower adjusted risk of escalation to ECMO (0.52, 95% CI 0.29 to 0.95) and LVAD implantation (0.28, 95% CI 0.08 to 0.94). CONCLUSIONS:The present study identified wide interhospital variation in the use of IABP and PVAD for AMI CS. Although the likelihood of escalation to ECMO or LVAD differed between hospital types, resource utilisation was similar.

Objective Postoperative atrial fibrillation (POAF) is the most common complication after coronary artery bypass grafting (CABG). This study explored the association of coronary sinus (CS) lactate levels with POAF following CABG. Methods This was a single-center prospective cohort study of patients without a history of atrial fibrillation undergoing isolated CABG (12/2024-2/2026). CS lactate was sampled before initiation of cardiopulmonary bypass (pre-CPB), prior to cross-clamp removal (pre-XC), and after cross-clamp removal (post-XC). POAF was defined as electrocardiogram documented atrial fibrillation treated with amiodarone during the index hospitalization. Restricted cubic splines were used to define a binary cutoff of post-XC CS lactate at >3 mmol/L. Multivariable logistic regression was performed to assess the association between CS lactate elevation and POAF. Results Of the 94 patients included, POAF occurred in 24 patients (26%). Patients with POAF were older and more commonly male but had similar median cardiopulmonary bypass time and cross-clamp duration. Patients with a post-XC lactate >3mmol/L (n=13/94, 14%) had significantly greater rates of POAF compared to those without lactate elevation (62% vs 20%, p=0.001). Following risk-adjustment, post-XC lactate >3 mmol/L was associated with a 11.7-fold increase in odds of POAF (p=0.010). Increasing age was also linked to greater likelihood of POAF. There was no significant association of POAF with CPB time, XC duration, pre-CPB CS lactate, or pre-XC CS lactate. Conclusions Elevated CS lactate after cardiac reperfusion was associated with increased likelihood of POAF. Our findings suggest that CS lactate may be used to guide POAF prophylaxis.

BACKGROUND:Although postoperative cardiac arrest is a well-studied complication of cardiac surgery, few guidelines exist regarding timing of surgery in preoperative cardiac arrest (pCA). We examined the association between delayed timing of operation and postoperative outcomes following cardiac surgery in a large cohort of pCA. METHODS:Adults with a diagnosis of pCA undergoing a cardiac operation were identified in the 2016-2020 National Inpatient Sample. Those requiring surgery within 24 hours fo cardiac arrest were excluded. Patients who underwent a cardiac procedure after 5 days of cardiopulmonary resuscitation were classified as Delayed (others: Early). Multivariable regression models were constructed to evaluate associations between delayed timing of surgery with in-hospital mortality, postoperative complications, hospitalization duration, and costs. RESULTS:Of an estimated 9,240 patients meeting study criteria, 4,860 (52.6%) received delayed cardiac surgery. Following entropy balancing, delayed surgery was significantly associated with decreased odds of in-hospital mortality (Adjusted Odds Ratio [AOR] 0.75, 95% Confidence Interval [CI] 0.58 - 0.97). However, delayed operation demonstrated greater odds of postoperative thromboembolic (AOR 1.44, 95% CI 1.02 - 2.04), and infectious (AOR 1.65, 95% CI 1.31 - 2.08) complications. Notably, delay did not alter odds of neurologic complication, and was linked to a decrement in per-day costs (β -$2,100, 95% CI -2,600 - -1,700). CONCLUSIONS:While preoperative cardiac arrest remains challenging, the present study demonstrates the safety profile of delaying cardiac operation among patients tolerating at least 24 hours of a delay to surgery. Future studies are needed to elucidate the factors associated with favorable outcomes in this population.

Background:Coronary artery bypass grafting (CABG) is traditionally performed though median sternotomy for multivessel coronary artery disease. Robotic CABG, a viable alternative, comprises less than 1% of CABG procedures in the United States despite its potential benefits. This study aimed to compare the trends and outcomes of conventional and robotic CABG by using a contemporary national cohort. Methods:A retrospective study was conducted using the 2016 to 2020 Nationwide Readmissions Database (NRD). Adult patients (aged ≥18 years) who underwent single-vessel CABG were identified using International Classification of Diseases, 10th revision procedure codes. Patients were categorized into robotic (totally endoscopic or robotic-assisted) and conventional CABG cohorts. Outcomes evaluated included in-hospital mortality, major adverse events (MAEs), length of stay, hospitalization costs, nonhome discharge, and 30-day readmissions. Results:Among 21,870 patients, 3433 (15.7%) underwent robotic CABG. The use of robotic CABG increased modestly over the study period. Patients who underwent robotic CABG had lower in-hospital mortality (0.4% vs 1.7%; P < .001) and MAEs (11.4% vs 18.9%; P < .001) compared with conventional CABG. Moreover, the robotic CABG cohort was associated with shorter length of stay and reduced hospitalization costs. After adjusting for baseline characteristics, robotic CABG showed lower odds of in-hospital mortality (adjusted odds ratio, 0.35; 95% CI, 0.15-0.84; P = .019) and MAEs (adjusted odds ratio, 0.72; 95% CI, 0.59-0.88; P = .001). Conclusions:Robotic CABG is associated with reduced in-hospital mortality, complications, LOS, and hospitalization costs compared with conventional CABG. Despite these benefits, its adoption remains limited, potentially because of the steep learning curve and resource requirements. Further efforts to overcome these barriers could enhance the adoption of robotic CABG and improve patient outcomes.

BACKGROUND:Failure to rescue has been increasingly used as a surgical quality metric, although implementation with complication-agnostic risk models may disproportionately penalize centers that care for high-risk patients. We used a nationally representative database to assess the impact of complication-sensitive risk models on hospital benchmarking for failure to rescue. METHODS:All adults undergoing elective coronary artery bypass grafting, aortic/mitral valve replacement, or esophageal/pancreatic/large bowel resection were identified within the 2019 Nationwide Readmissions Database. Two hierarchical logistic regressions (model 1: complication-agnostic; model 2: complication-sensitive) were developed to evaluate risk-adjusted rates of failure to rescue at each center. Patient characteristics (demographics, comorbidities) were incorporated as fixed effects in both models. Model 2 also included adjustment for the occurrence and identity of each complication. Hospitals were subsequently grouped into quintiles of failure to rescue using each model. RESULTS:Approximately 296,907 patients at 1,034 hospitals met inclusion criteria. Overall mortality, complication, and failure to rescue rates were 1.1%, 4.8%, and 17.8%, respectively. Centers in the highest quintile of failure to rescue for model 1 more frequently managed patients who developed cardiac arrest (0.9 vs 0.7%, P = .003) or acute kidney injury requiring dialysis (0.6 vs 0.4%, P = .017). In contrast, the rates of all complications except sepsis (2.7 vs 2.3%, P = .035) were comparable between centers in the top quintile and others, when using model 2. Overall, ∼30% of hospitals were reclassified into different quintiles with the complication-sensitive model. CONCLUSION:This study suggests that complication-agnostic models disproportionately penalize centers caring for patients who develop severe complications, which can be mitigated with complication-sensitive models.

Background: Despite increasing utilization and survival benefit over the last decade, extracorporeal membrane oxygenation (ECMO) remains resource-intensive with significant complications and rehospitalization risk. We thus utilized machine learning (ML) to develop prediction models for 90-day nonelective readmission following ECMO. Methods: All adult patients receiving ECMO who survived index hospitalization were tabulated from the 2016-2020 Nationwide Readmissions Database. Extreme Gradient Boosting (XGBoost) models were developed to identify features associated with readmission following ECMO. Area under the receiver operating characteristic (AUROC), mean Average Precision (mAP), and the Brier score were calculated to estimate model performance relative to logistic regression (LR). Shapley Additive Explanation summary (SHAP) plots evaluated the relative impact of each factor on the model. An additional sensitivity analysis solely included patient comorbidities and indication for ECMO as potential model covariates. Results: Of similar to 22,947 patients, 4495 (19.6 %) were readmitted nonelectively within 90 days. The XGBoost model exhibited superior discrimination (AUROC 0.64 vs 0.49), classification accuracy (mAP 0.30 vs 0.20) and calibration (Brier score 0.154 vs 0.165, all P < 0.001) in predicting readmission compared to LR. SHAP plots identified duration of index hospitalization, undergoing heart/lung transplantation, and Medicare insurance to be associated with increased odds of readmission. Upon sub-analysis, XGBoost demonstrated superior disclination compared to LR (AUROC 0.61 vs 0.60, P < 0.05). Chronic liver disease and frailty were linked with increased odds of nonelective readmission. Conclusions: ML outperformed LR in predicting readmission following ECMO. Future work is needed to identify other factors linked with readmission and further optimize post-ECMO care among this cohort.

Objective. To assess perioperative and readmission outcomes of patients undergoing head and neck cancer (HNCA) surgery at safety-net hospitals (SNHs) in a modern cohort. Study Design. Retrospective cohort study. Setting. Nationwide Readmissions Database (NRD), 2010 to 2019. Methods. All elective adult (>= 18 years) admissions involving HNCA resection were identified from the NRD. To calculate safety-net burden, the proportion of Medicaid or uninsured patients admitted to each hospital for any indication was tabulated annually, with centers in the highest quartile defined as SNHs. To perform risk adjustment in assessing perioperative and readmission outcomes, multivariable regression models were developed. Results. Of an estimated 133,018 head and neck surgical patients, 26.5% (n = 35,268) received treatment at a SNH. Utilization of SNHs increased over the decade-long study period, with 29.8% of individuals treated at these sites in 2019. After multivariable adjustment, several patient factors were noted to be associated with SNHs, including younger age, lower comorbidity burden, and income within the lowest quartile. Although incidence of adverse events decreased at both SNHs and non-SNHs during the study period, treatment at SNHs remained associated with these events after risk adjustment (adjusted odds ratio: 1.17, 95% confidence interval: 1.08-1.28, P < .001). Conclusion. SNHs continue to provide valuable specialty care to underserved populations, often with limited financial resources. Despite promising results from prior decades demonstrating comparable perioperative outcomes, the present study noted increased adverse events following HNCA surgery at these sites. Such findings underscore the need for continued advocacy to secure necessary funding for these centers.

BackgroundExpedited discharge following esophagectomy is controversial due to concerns for higher readmissions and financial burden. The present study aimed to evaluate the association of expedited discharge with hospitalization costs and unplanned readmissions following esophagectomy for malignant lesions.MethodsAdults undergoing elective esophagectomy for cancer were identified in the 2014-2019 Nationwide Readmissions Database. Patients discharged by postoperative day 7 were considered Expedited and others as Routine. Patients who did not survive to discharge or had major perioperative complications were excluded. Multivariable regression models were constructed to assess association of expedited discharge with index hospitalization costs as well as 30- and 90-day non-elective readmissions.ResultsOf 9,886 patients who met study criteria, 34.6% comprised the Expedited cohort. After adjustment, female sex (adjusted odds ratio [AOR] 0.71, p = 0.001) and increasing Elixhauser Comorbidity Index (AOR 0.88/point, p<0.001) were associated with lower odds of expedited discharge, while laparoscopic (AOR 1.63, p<0.001, Ref: open) and robotic (AOR 1.67, p = 0.003, Ref: open) approach were linked to greater likelihood. Patients at centers in the highest-tertile of minimally invasive esophagectomy volume had increased odds of expedited discharge (AOR 1.52, p = 0.025, Ref: lowest-tertile). On multivariable analysis, expedited discharge was independently associated with an $8,300 reduction in hospitalization costs. Notably, expedited discharge was associated with similar odds of 30-day (AOR 1.10, p = 0.40) and 90-day (AOR 0.90, p = 0.70) unplanned readmissions.ConclusionExpedited discharge after esophagectomy was associated with decreased costs and unaltered readmissions. Prospective studies are necessary to robustly evaluate whether expedited discharge is appropriate for select patients undergoing esophagectomy.

Background: Neoadjuvant therapy is being increasingly used for patients with pancreatic cancer. The role of adjuvant therapy in these patients is unclear. The purpose of this study was to identify clinical and pathologic characteristics that are associated with longer overall survival in patients with pancreatic cancer who receive adjuvant therapy after neoadjuvant therapy. Methods: This study was conducted using multi -institutional data. All patients underwent surgery after at least 1 cycle of neoadjuvant therapy for pancreatic cancer. Patients who died within 3 months after surgery and were known to have distant metastasis or macroscopic residual disease were excluded. Mann -Whitney U test, c 2 analysis, Kaplan -Meier plot, and univariate and multivariate Cox regression analysis were performed as statistical analyses. Results: In the present study, 529 patients with resected pancreatic cancer after neoadjuvant therapy were reviewed. For neoadjuvant therapy, 177 (33.5%) patients received neoadjuvant chemotherapy, and 352 (66.5%) patients received neoadjuvant chemoradiotherapy. The median duration of neoadjuvant therapy was 7.0 months (interquartile range, 5.0-8.7). Patients were followed for a median of 23.0 months after surgery. Adjuvant therapy was administered to 297 (56.1%) patients and was not associated with longer overall survival for the entire cohort (24 vs 22 months, P = .31). Interaction analysis showed that adjuvant therapy was associated with longer overall survival in patients who received less than 4 months neoadjuvant therapy (hazard ratio 0.40; 95% con fidence interval 0.17-0.95; P =.03) or who had microscopic margin positive surgical resections (hazard ratio 0.56; 95% con fidence interval 0.33-0.93; P = .03). Conclusion: In this retrospective study, there was a survival bene fit associated with adjuvant therapy for patients who received less than 4 months of neoadjuvant therapy or had microscopic positive margins. (c) 2024 Elsevier Inc. All rights reserved.

OBJECTIVE:To create a novel comorbidity score tailored for surgical database research. BACKGROUND:Despite their use in surgical research, the Elixhauser (ECI) and Charlson (CCI) Comorbidity Indices were developed nearly 4 decades ago utilizing primarily nonsurgical cohorts. METHODS:Adults undergoing 62 operations across 14 specialties were queried from the 2019 National Inpatient Sample (NIS), using the International Classification of Diseases, 10th Revision codes. International Classification of Diseases, 10th Revision codes for chronic diseases were sorted into Clinical Classifications Software Refined groups. Clinical Classifications Software Refined with non-zero feature importance across 4 machine learning algorithms predicting in-hospital mortality were used for logistic regression; resultant coefficients were used to calculate the Comorbid Operative Risk Evaluation (CORE) score based on previously validated methodology. Areas under the receiver operating characteristic with 95% CIs were used to compare model performance in predicting in-hospital mortality for the CORE score, ECI, and CCI. Validation was performed using the 2016-2018 NIS, combined 2018-2019 Florida and New York State Inpatient Databases (SID), and 2016-2022 institutional data. RESULTS:A total of 699,155 records from the 2019 NIS were used for model development. The CORE score better predicted in-hospital mortality compared with the ECI within the NIS (0.90, 95% CI: 0.90-0.90 vs 0.84, 95% CI: 0.84-0.84), SID (0.91, 95% CI: 0.90-0.91 vs 0.86, 95% CI: 0.86-0.87), and institutional (0.88, 95% CI: 0.87-0.89 vs 0.84, 95% CI: 0.83-0.85) databases (all P < 0.001). Likewise, it outperformed the CCI for the NIS (0.76, 95% CI: 0.76-0.76), SID (0.78, 95% CI: 0.77-0.78), and institutional (0.62, 95% CI: 0.60-0.64) cohorts (all P < 0.001). CONCLUSIONS:The CORE score may better predict in-hospital mortality after surgery due to comorbid diseases in outcome-based research.

Objective: Although national efforts have aimed to improve the safety of inpatient operations, income-based inequities in surgical outcomes persist, and the evolution of such disparities has not been examined in the contemporary setting. We sought to examine the association of community-level household income with acute outcomes of cardiac procedures over the past decade. Methods: All adult hospitalizations for elective coronary artery bypass grafting/ valve operations were tabulated from the 2010-2020 Nationwide Readmissions Database. Patients were stratified fi ed into quartiles of income, with records in the 76th to 100th percentile designated as highest and those in the 0 to 25th percentile as lowest. To evaluate the change in adjusted risk of in-hospital mortality, complications, and readmission over the study period, estimates were generated for each income level and year. Results: Of approximately 1,848,755 hospitalizations, 406,216 patients (22.0%) % ) were classified fi ed as highest income and 451,988 patients (24.4%) % ) were classified fi ed as lowest income. After risk adjustment, lowest income remained associated with greater likelihood of in-hospital mortality (adjusted odds ratio, 1.61, 95% % CI, 1.51-1.72), any postoperative complication (adjusted odds ratio, 1.19, CI, 1.151.22), and nonelective readmission within 30 days (adjusted odds ratio, 1.07, CI, 1.05-1.10). Overall adjusted risk of mortality, complications, and nonelective readmission decreased for both groups from 2010 to 2020 (P P < .001). Further, the difference in risk of mortality between patients of lowest and highest income decreased by 0.2%, % , whereas the difference in risk of major complications declined by 0.5% % (both P < .001). Conclusions: Although overall in-hospital mortality and complication rates have declined, low-income patients continue to face greater postoperative risk. Novel interventions are needed to address continued income-based disparities and ensure equitable surgical outcomes. (JTCVS Open 2024;20:89-100)

Introduction: Accurate prediction of complications often informs shared decision-making. Derived over 10 years ago to enhance prediction of intra/ post-operative myocardial infarction and cardiac arrest (MI/CA), the Gupta score has been criticized for unreliable calibration and inclusion of a wide spectrum of unrelated operations. In the present study, we developed a novel machine learning (ML) model to estimate perioperative risk of MI/CA Methods: Patients undergoing major operations were identified from the 2016-2020 ACS-NSQIP. The Gupta score was calculated for each patient, and a novel ML model was developed to predict MI/CA using ACS NSQIP-provided data fields as covariates. Discrimination (C-statistic) and calibration (Brier score) of the ML model were compared to the existing Gupta score within the entire cohort and across operative subgroups. Results: Of 2,473,487 patients included for analysis, 25,177 (1.0%) experienced MI/CA (55.2% MI, 39.1% CA, 5.6% MI and CA). The ML model, which was fit using a randomly selected training cohort, exhibited higher discrimination within the testing dataset compared to the Gupta score (Cstatistic 0.84 vs 0.80, p < 0.001). Furthermore, the ML model had significantly better calibration in the entire cohort (Brier score 0.0097 vs 0.0100). Model performance was markedly improved among patients undergoing thoracic, aortic, peripheral vascular and foregut surgery. Conclusions: The present ML model outperformed the Gupta score in the prognostication of MI/CA across a heterogenous range of operations. Given the growing integration of ML into healthcare, such models may be readily incorporated into clinical practice and guide benchmarking efforts.

BackgroundAlthough early discharge after colectomy has garnered significant interest, contemporary, large-scale analyses are lacking.ObjectiveThe present study utilized a national cohort of patients undergoing colectomy to examine costs and readmissions following early discharge.MethodsAll adults undergoing elective colectomy for primary colon cancer were identified in the 2016-2019 Nationwide Readmissions Database. Patients with perioperative complications or prolonged length of stay (>8 days) were excluded to enhance cohort homogeneity. Patients discharged by postoperative day 3 were classified as Early, and others as Routine. Entropy balancing and multivariable regression were used to assess the risk-adjusted association of early discharge with costs and non-elective readmissions. Importantly, we compared 90-day stroke rates to examine whether our results were influenced by preferential early discharge of healthier patients.ResultsOf an estimated 153,996 patients, 45.5% comprised the Early cohort. Compared to Routine, the Early cohort was younger and more commonly male. Patients in the Early group more commonly underwent left-sided colectomy and laparoscopic operations. Following multivariable adjustment, expedited discharge was associated with a $4,500 reduction in costs as well as lower 30-day (adjusted odds ratio [AOR] 0.74, p<0.001) and 90-day non-elective readmissions (AOR 0.74, p<0.001). However, among those readmitted within 90 days, Early patients were more commonly readmitted for gastrointestinal conditions (45.8 vs 36.4%, p<0.001). Importantly, both cohorts had comparable 90-day stroke rates (2.2 vs 2.1%, p = 0.80).ConclusionsThe present work represents the largest analysis of early discharge following colectomy for cancer and supports its relative safety and cost-effectiveness.

BACKGROUND Using a nationally representative database, the present study evaluated the degree of center -level variation in the cost of transcatheter aortic valve replacement (TAVR). METHODS All adults undergoing elective, isolated TAVR were identified in the 2016 to 2018 Nationwide Readmissions Database. Multilevel mixed -effects models were used to identify patient and hospital characteristics associated with hospitalization costs. The random intercept for each hospital was generated and considered to be the baseline cost attributable to care at each center. Hospitals in the highest decile of baseline costs were classified as high -cost hospitals. The association of high -cost hospital status with in -hospital mortality and perioperative complications was subsequently assessed. RESULTS An estimated 119,492 patients, with a mean age of 80 years and a 45.9% prevalence of female sex, met the study criteria. Analysis of random intercepts indicated that 54.3% of variability in costs was attributable to interhospital differences rather than patient factors. Perioperative respiratory failure, neurologic complications, and acute kidney injury were associated with increased episodic expenditure but did not explain the observed center -level variation. The baseline cost associated with each hospital ranged from -$26,000 to $162,000. Notably, high -cost hospital status was not linked to annual TAVR caseload or to odds of mortality (P = .83), acute kidney injury (P = .18), respiratory failure (P = .32), or neurologic complications (P = .55). CONCLUSIONS The present analysis identified significant variation in the cost of TAVR, which was largely attributable to center -level rather than patient factors. Hospital TAVR volume and occurrence of complications were not drivers of the observed variation.

OBJECTIVE:The aim of this study was to develop a novel machine learning model to predict clinically relevant postoperative pancreatic fistula (CR-POPF) following pancreaticoduodenectomy (PD). BACKGROUND:Accurate prognostication of CR-POPF may allow for risk stratification and adaptive treatment strategies for potential PD candidates. However, antecedent models, such as the modified Fistula Risk Score (mFRS), are limited by poor discrimination and calibration. METHODS:All records entailing PD within the 2014 to 2018 American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) were identified. In addition, patients undergoing PD at our institution between 2013 and 2021 were queried from our local data repository. An eXtreme Gradient Boosting (XGBoost) model was developed to estimate the risk of CR-POPF using data from the ACS NSQIP and evaluated using institutional data. Model discrimination was estimated using the area under the receiver operating characteristic (AUROC) and area under the precision recall curve (AUPRC). RESULTS:Overall, 12,281 and 445 patients undergoing PD were identified within the 2014 to 2018 ACS NSQIP and our institutional registry, respectively. Application of the XGBoost and mFRS scores to the internal validation dataset revealed that the former model had significantly greater AUROC (0.72 vs 0.68, P <0.001) and AUPRC (0.22 vs 0.18, P <0.001). Within the external validation dataset, the XGBoost model remained superior to the mFRS with an AUROC of 0.79 (95% CI: 0.74-0.84) versus 0.75 (95% CI: 0.70-0.80, P <0.001). In addition, AUPRC was higher for the XGBoost model, compared with the mFRS. CONCLUSION:Our novel machine learning model consistently outperformed the previously validated mFRS within internal and external validation cohorts, thereby demonstrating its generalizability and utility for enhancing prediction of CR-POPF.

BACKGROUND:Although financial toxicity, defined as the harmful financial burden experienced by patients undergoing cancer treatment, has been of growing interest, data in thoracic oncology are lacking. This study aimed to examine the risk of financial toxicity among patients undergoing surgical resection of thoracic malignant diseases. METHODS:Adults undergoing lobectomy, pneumonectomy, or esophagectomy for cancer were identified in the 2012 to 2021 National Inpatient Sample. Risk of financial toxicity was defined as health expenditure (total hospitalization costs for the uninsured and maximum out-of-pocket costs for the insured) exceeding 40% of postsubsistence income. Multivariable logistic regressions were used to identify factors associated with financial toxicity risk. RESULTS:Of 384,340 patients, 69.5% had government-funded insurance, 27.2% had private insurance, and 1.0% were uninsured. Compared with patients with insurance, uninsured patients were more commonly Black and Hispanic and less commonly electively admitted. Mortality, complications, length of stay, and costs were comparable regardless of insurance status. Approximately 68.9% of uninsured and 17.3% of insured patients were at risk of financial toxicity, and the incidence of financial toxicity remained stable over time. After risk adjustment, complications were associated with a greater than 2-fold increased risk of financial toxicity among uninsured patients (adjusted odds ratio, 2.21; 95% CI, 1.38-3.55). Among the insured patients, Black, Hispanic, and publicly insured patients demonstrated a greater risk of financial toxicity, while patients undergoing minimally invasive operations and receiving care at metropolitan hospitals exhibited a lower risk of financial toxicity. CONCLUSIONS:Concordant with previous work examining financial toxicity in abdominal oncologic surgery, thoracic surgery demonstrates a comparable burden of financial toxicity. Referral policies and care subsidization may be considered in patients at risk for financial toxicity who are undergoing resections for thoracic malignant diseases.