Search

Whereas the association between surgical volume and outcomes has been well established, the potential impact of specialized pediatric centers on outcomes of cardiac operations for adults with congenital heart disease has not been elucidated.The 2010-2017 Nationwide Readmissions Database was queried to identify all adults with congenital heart disease. High-volume centers were designated the highest tertile of operative case volume annually for both pediatric and adult cardiac operations. Multivariable regression models adjusting for demographic and clinical characteristics were used to evaluate adjusted odds ratios for select outcomes.Of an estimated 52 357 hospitalizations meeting inclusion criteria, 6074 (11.7%) received an operation at a pediatric high-volume center (pHVC) and 45 652 (87.2%) at an adult high-volume center (aHVC). Compared with an aHVC, patients at a pHVC were on average younger, had a similar Elixhauser Comorbidity Index, and underwent higher risk operations. They more commonly carried private insurance and were categorized within the top income quartile. On multivariable analysis, operations at a pHVC were associated with reduced odds of perioperative complications (adjusted odds ratio [AOR], 0.85; 95% CI, 0.72-0.99), nonhome discharge (AOR, 0.64; 95% CI, 0.55-0.73), and 90-day emergent readmissions (AOR, 0.73; 95% CI, 0.60-0.89) but similar risk of death (AOR, 0.74; 95% CI, 0.43-1.28).Compared with high-volume hospitals for adult cardiac operations, congenital heart disease operations at high-volume pediatric cardiac centers were associated with reduced odds of complications, nonhome discharges, and urgent readmissions. Our findings may better inform appropriate referral of this cohort of complex patients and regionalization of their care.

Author(s): Seo, Young-Ji; Sareh, Sohail; Hadaya, Joseph; Sanaiha, Yas; Ziaeian, Boback; Shemin, Richard J; Benharash, Peyman | Abstract: BackgroundLack of consensus remains about factors that may be associated with high resource use (HRU) in adult cardiac surgical patients. This study aimed to identify patient-related, hospital, and perioperative characteristics associated with HRU admissions involving elective cardiac operations.MethodsData from the National Inpatient Sample was used to identify patients who underwent coronary artery bypass graft, valve replacement, and valve repair operations between 2005 and 2016. Admissions with HRU were defined as those in the highest decile for total hospital costs. Multivariable regressions were used to identify factors associated with HRU.ResultsAn estimated 1,750,253 hospitalizations coded for elective cardiac operations. The median hospitalization cost was $34,700 (interquartile range, $26,800- to $47,100), with the HRU (Nn= 175,025) cutoff at $66,029. Although HRU patients comprised 10% of admissions, they accounted for 25% of cumulative costs. On multivariable regression, patient-related characteristics predictive of HRU included female sex, older age, higher comorbidity burden, non-White race, and highest income quartile. Hospital factors associated with HRU were low-volume hospitals for both coronary artery bypass graft and valvular operations. Among postoperative outcomes, mortality, infectious complications, extracorporeal membrane oxygenation use, and hospitalization for more than 8 days were associated with greater odds of HRU.ConclusionsIn this nationwide study of elective cardiac surgical patients, several important patient-related and hospital factors, including patients' race, comorbidities, postoperative infectious complications, and low hospital operative volume were identified as predictors of HRU. These highly predictive factors may be used for benchmarking purposes and improvement in surgical planning.

Frailty has been widely recognized as a predictor of postoperative outcomes. Given the paucity of standardized frailty measurements in thoracic procedures, this study aimed to determine the impact of coding-based frailty on clinical outcomes and resource use after anatomic lung resection.All adults undergoing elective, anatomic lung resections (segmentectomy, lobectomy, pneumonectomy) from 2005 to 2014 were identified using the National Inpatient Sample. Patients were categorized as either frail or nonfrail on the basis of the presence of any frailty-defining diagnoses defined by the Johns Hopkins Adjusted Clinical Groups. Multivariable models were used to assess the independent association of frailty with in-hospital mortality, nonhome discharge, complications, duration of stay, and costs.Of an estimated 366,357 hospitalizations for elective lung resection during the study period, 4.4% were in frail patients. Patients who underwent pneumonectomy or were treated at low-volume hospitals were more commonly frail. Relative to nonfrail patients, frailty was associated with increased unadjusted mortality (9.1% vs 1.7%; P.001) and nonhome discharge (44.7% vs 10.5%; P.001). Frail patients had 3.47 increased adjusted odds of mortality across resection types (95% confidence interval, 2.94 to 4.09). Frailty conferred the greatest increase in mortality, complications, and resource use after pneumonectomy relative to lobectomy or segmentectomy, although significant differences were evident for all 3 operations.Frailty exhibits a strong association with inferior clinical outcomes and increased resource use after elective lung resection, particularly pneumonectomy. This readily available tool may improve preoperative risk assessment and allow for better selection of treatment modalities for frail patients with pulmonary disorders.

Although not formalized into current risk assessment models, frailty has been associated with negative postoperative outcomes in many specialties. Using administrative coding, we evaluated the impact of frailty on in-hospital death, complications, and resource use in a nationally representative cohort of patients undergoing isolated coronary artery bypass grafting (CABG).Patients aged 18 years and older who underwent isolated CABG across the United States were identified using the 2005 to 2016 National Inpatient Sample. Frailty was defined using a derivative of the validated Johns Hopkins Adjusted Clinical Groups frailty-defining diagnosis indicator. Mortality, length of stay, inflation-adjusted costs, and postoperative complications were evaluated using multilevel multivariable regression.Of an estimated 2,137,618 patients undergoing isolated CABG, 85,879 (4.0%) were considered frail. The proportion of frail patients increased over the study period (nonparametric test for trend P = .002), while annual mortality rates declined (nonparametric test for trend P.001). Frail patients were older (68.9 ± 10.7 years vs 65.0 ± 10.6 years, P.001), and more commonly female (32.8% vs 26.2%, P.001). After adjustment, frailty was associated with increased odds of in-hospital death (adjusted odds ratio [AOR], 2.49; 95% confidence interval [CI], 2.30-2.70; P.001), major complications (AOR, 2.55; 95% CI, 2.39-2.71; P.001), increased length of stay (AOR, 1.40; 95% CI, 1.09-2.11; P.001), and costs (AOR, 1.03; 95% CI, 1.02-1.07; P.001).Frailty, as identified by administrative coding, serves as a strong independent predictor of death and complications after CABG. Incorporation of frailty into risk models may aid in counseling patients about operative risk and benchmarking outcomes.

Background The objective of this study was to characterize practical use trends and outcomes for intraaortic balloon pump (IABP) and percutaneous left ventricular assist device (pVAD) use in cardiogenic shock at a national level. Methods An analysis of all adult patients admitted nonelectively for cardiogenic shock from January 2008 through December 2017 was performed using the National Inpatient Sample. Trends of inpatient IABP and pVAD use were analyzed using survey-weighted estimates and the modified Cochran-Armitage test for significance. Multivariable regression models and inverse probability of treatment weights were used to perform risk-adjusted analyses of pVAD mortality, a composite of adverse events (AE), and resource use, with IABP as reference. Results Of an estimated 774,310 patients admitted with cardiogenic shock, 143,051 received a device: IABP, 127,792 (16.5%); or pVAD, 15,259 (2.0%). IABP use decreased (23.8% to 12.7%; P for trend Conclusions Over the study period, the rate of pVAD use for cardiogenic shock significantly increased. Compared with IABP, pVAD use was associated with increased mortality, higher costs, and several AEs. Multi-institutional clinical trials with rigorous inclusion criteria are warranted to evaluate the clinical utility of pVADs in the modern era.

Autoimmune connective tissue diseases (CTDs) are associated with accelerated atherosclerosis and inflammation, while often requiring immunosuppression. Large-scale outcomes of coronary artery bypass graft (CABG) surgery in this population have not been reported thus far. This study characterized trends in use of CABG in patients with CTDs and the impact of the disease on mortality, in-hospital complications, length of stay, and costs.The 2005 to 2015 National Inpatient Sample was used to identify all adult patients undergoing isolated CABG. The CTDs cohort included rheumatoid arthritis, lupus erythematosus, and antiphospholipid syndrome (APLS), among others. Hierarchical multivariable logistic models were used to calculate the independent impact of CTDs on clinical outcomes and costs.Of an estimated 2,101,591 patients, 41,567 (1.8%) were diagnosed with CTDs (rheumatoid arthritis, 58%; systemic lupus erythematosus, 12%; APLS, 11%) Although the overall annual use of CABG decreased, the proportion of patients with CTDs receiving the operation significantly increased. After adjusting for patient and hospital characteristics, CTDs were not associated with increased mortality (adjusted odds ratio [AOR], 0.91; P = .34) but were protective against cardiovascular (AOR, 0.92; P.003), neurologic (AOR, 0.81; P = .01), and infectious (AOR, 0.80; P = .01) complications. The diagnosis of CTDs was also predictive of reduced length of hospital stay (β-coefficient = -0.40; P.001) and costs (β-coefficient, -$1200; P = .01). On subgroup analysis patients with APLS had significantly increased odds of mortality (AOR, 1.5) and increased renal (AOR, 1.3), infectious (AOR, 1.7), and thromboembolic (AOR, 4.3) complications (all P.05).CABG in patients with CTDs provides acceptable outcomes and paradoxically improved resource use. However CABG in patients with APLS warrants careful consideration given inferior outcomes.

Background Despite advances in surgical technique and perioperative management, pneumonectomy remains associated with significant morbidity and mortality. The purpose of this study was to examine the impact of annual institutional volume of anatomic lung resections on outcomes after elective pneumonectomy. Methods We evaluated all patients who underwent elective pneumonectomy from 2005 to 2014 in the National Inpatient Sample. Patients less than 18 years of age, or with trauma-related diagnoses, mesothelioma, or a nonelective admission were excluded. Hospitals were divided into volume quartiles based on annual institutional anatomic lung resection caseload. We studied the effect of institutional volume on inhospital mortality, complications, and failure to rescue, as well as costs and length of stay. Results During the study period, an estimated 22,739 patients underwent pneumonectomy, with a reduction in national mortality from 7.9% to 5.5% (P trend = .045). Compared with the highest volume centers, operations performed at the lowest volume hospitals were associated with 1.74 increased odds of mortality (95% confidence interval, 1.14 to 2.66). Despite similar odds of postoperative complications, low volume hospital status was associated with increased failure to rescue rates (18.3% vs 12.7%, P = .024) and adjusted odds of mortality (1.70; 95% confidence interval, 1.09 to 2.64) after any complication. Conclusions High volume hospital status is strongly associated with reduced mortality and failure to rescue rates after pneumonectomy. Efforts to centralize care or disseminate best practices may lead to improved national outcomes for this high-risk procedure.

Background Clostridium difficile infection (CDI) has been associated with morbidity and mortality after cardiac operations. The present study examined incidence, predictors, and impact of CDI on inpatient mortality and resource utilization. Methods An analysis of adult patients undergoing elective coronary artery bypass grafting or valvular operations from 2005 to 2016 was performed using the National Inpatient Sample. Trends in CDI were assessed using a modified Cochran-Armitage analysis. Multivariable multilevel regressions were used to identify predictors of CDI, and propensity-matched pairs were generated using Mahalanobis 1-to-1 matching to compare mortality, length of stay, and costs of CDI patients with the non-CDI cohort. Results The overall rate of CDI for an estimated 2,026,267 patients who underwent elective major cardiac surgery was 0.5% with no change in incidence (P for trend = .99). Predictors of CDI included advanced age (≥65 y; adjusted odds ratio [AOR], 1.88; 95% confidence interval [CI], 1.58-2.24), female gender (AOR, 1.29; 95% CI, 1.15-1.44), heart failure (AOR, 1.57; 95% CI, 1.40-1.76), and combined coronary artery bypass grafting/valve operations (AOR, 1.60; 95% CI, 1.24-2.08). Neither region nor bed size was associated with CDI. In contrast CDI mortality was lower at teaching hospitals compared with rural hospitals. Among matched pairs CDI was independently associated with higher mortality, length of stay, and Gross Domestic Product-adjusted costs. Conclusions CDI occurs in less than 1% of all elective, major cardiac operations. Patient predictors included advanced age, female gender, and several chronic comorbidities. Teaching institutions had the highest odds of CDI but lowest odds of case fatality. Further investigation of factors contributing to CDI is warranted to disseminate institutional best practices.

Reducing inpatient readmissions is a national priority for improving healthcare quality and decreasing costs. Previous studies have shown that readmissions after surgical aortic valve replacement are frequent and contribute to increased healthcare costs, yet no studies have analyzed risk factors for readmission.The Nationwide Readmissions Database was used to identify adult patients undergoing surgical aortic valve replacement from 2010 to 2015. Incidence, patient characteristics, causes, resource utilization, and predictors of 30-day readmission were determined. International Classification of Diseases codes were used to capture surgical aortic valve replacement.Among 136,051 patients, 18,631 (13.7%) were readmitted within 30 days of discharge. Readmitted patients were more commonly women (47.4% vs 41.6%; P.001) and were older (70.4 years of age vs 68.3 years of age; P.001), with higher Elixhauser comorbidity index (5.4 vs 4.8; P.001), rates of postoperative complications (44.0% vs 37.3%; P.001), and greater length of stay (10.9 days vs 8.5 days; P.001). The mean cost of 1 readmission episode was $13,426. On multivariable analysis, significant predictors of readmission were female sex, age greater than 75 years, atrial fibrillation, chronic kidney and liver disease, and lower surgical aortic valve replacement hospital volume. A total of 49.1% of readmissions were related to cardiac causes, with heart failure (13.2%) and arrhythmia (12.5%) being the most common.Using a national inpatient database, we found readmission after surgical aortic valve replacement to be common and resource-intensive. Enhanced management of comorbidities and targeted postdischarge interventions for patients at high risk of readmission may help decrease healthcare utilization.

In the current era of value-based health care delivery, an understanding of patient populations at greatest risk for mortality, complications, and readmissions after thoracic endovascular aortic repair (TEVAR) is warranted. Thus, the present study aimed to evaluate outcomes after TEVAR for patients with varying degrees of renal dysfunction.All patients who underwent TEVAR from 2010 to 2015 in the Nationwide Readmissions Database were identified. These patients were further stratified into four groups: no chronic kidney disease (NCKD), chronic kidney disease (CKD) stages 1 to 3 (CKD1-3), CKD 4 to 5 (CKD4-5), and end-stage renal disease (ESRD) requiring dialysis. Multivariable regression analysis was used to study index mortality, early (30 days) and intermediate (31-90 days) readmissions, costs, and length of stay. Kaplan-Meier analyses were performed to compare readmission performance among all four groups.An estimated 121,046 patients underwent TEVAR with 26,653 (22.1%) being elective. Patients with ESRD comprised 2.7% of elective and 5.4% of nonelective TEVAR operations. Patients with CKD4-5 (17.8%; P = .01) and with ESRD (21.1%; P.001), but not with CKD1-3 (14.1%; P = .12), had remarkably higher early readmission rate than the NCKD cohort (9.2%). Patients with ESRD had remarkably higher hospitalization costs than the NCKD group ($7456; 95% confidence interval, $2629-$12,283). Cardiovascular, infectious, and vascular complications were the most prevalent diagnoses on readmission, with no remarkable difference among the NCKD and CKD4-5/ESRD groups.Nearly 10% of all patients with TEVAR have evidence of chronic kidney disease of varying severity. Only patients with ESRD are at risk of substantially higher odds of mortality, readmissions, index length of stay, and costs compared with the non-CKD cohort.

Transcatheter aortic valve replacement (TAVR) is not widely used in patients with bicuspid aortic valve (BAV) disease and has not yet been studied in randomized clinical trials. We characterized the rate of use and outcomes of TAVR and surgical aortic valve replacement (SAVR) in patients with BAV.Adults with BAV stenosis receiving SAVR or TAVR procedures were abstracted from the 2012 to 2019 Nationwide Readmissions Database (NRD). Risk-adjusted analyses were performed with NRD-provided weights and inverse probability of treatment weights (IPTW) to examine the association of treatment strategy on inpatient mortality, complications, and hospitalization resource utilization. Nonelective readmissions within 90 days of discharge and reintervention at the first readmission were also examined.Of an estimated 56 331 patients with BAV requiring aortic valve replacement, 6.8% underwent TAVR. Unadjusted analysis demonstrated higher index hospitalization mortality for TAVR compared with SAVR. Upon risk adjustment using NRD-provided weights, the odds of pacemaker implantation remained significantly higher for TAVR patients compared with SAVR, with no significant difference in mortality. When NRD-provided survey weights were applied, TAVR had higher rates of 90-day readmission. Adjustment with inverse probability of treatment weights resolved these differences between the 2 groups. Regardless of the risk-adjustment method, the odds of reintervention were consistently higher among BAV TAVR patients compared with SAVR.The present analysis demonstrates comparable in-hospital mortality and morbidity for TAVR and SAVR patients in the moderate-risk era. With increasing TAVR use in BAV, surgeons must further refine selection criteria with consideration of concomitant aortopathy and implications of reintervention.

Optimization of value, or quality relative to costs, has garnered significant attention in the United States. We aimed to characterize center-level variation in costs and quality after pulmonary lobectomy using a national cohort.Adults undergoing elective pulmonary lobectomy were identified in the 2016 to 2018 Nationwide Readmissions Database. Quality was defined by the absence of major adverse outcomes including respiratory failure, acute kidney injury, reoperation, and death. Risk-adjusted adverse outcome rates and costs were studied for institutions performing greater than or equal to 10 operations annually. Using observed-to-expected (O/E) ratios, high-value hospitals were defined as those with an O/E ratio less than 1 for costs and O/E ratio less than 1 for quality, while low-value hospitals were defined by the converse.Among 95 446 patients managed at 565 hospitals annually, the median center-level cost for lobectomy was $22 000 (interquartile range, $18 000-$27 000), while the median adverse outcome rate was 14.3% (interquartile range, 8.3%-23.1%). Centers with an O/E ratio less than 1 for adverse events exhibited a $2200/case reduction in risk-adjusted costs. Using O/E ratios, 35.2% of centers were classified as high value while 18.6% were low value. Compared with low-value centers, high-value centers treated older patients (67.1 years of age vs 65.5 years of age; P.001) with greater comorbidities (Elixhauser Comorbidity Index 3.7 vs 2.9; P.001) but had greater annual lobectomy volume (40 cases vs 30 cases; P = .001) and were more commonly teaching hospitals.Significant variation in costs and quality persists for lobectomy at the national level. Although high-value programs operated on patients at greater surgical risk, they had reduced complications and costs. Our findings suggest the need for dissemination of quality improvement and cost reduction practices.

BACKGROUND Efforts to reduce postoperative infections have garnered national attention, leading to practice guidelines for cardiac surgical perioperative care. The present study characterized the impact of health care-acquired infection (HAI) on index hospitalization costs and postdischarge health care utilization. METHODS Adults undergoing elective coronary artery bypass graft surgery (CABG) or valve operations were identified in the 2016 to 2018 Nationwide Readmissions Database. Infections were categorized into bloodstream, gastrointestinal, pulmonary, surgical site, or urinary tract infections. Generalized linear or flexible hazard models were used to assess associations between infections and outcomes. Observed-to-expected ratios were generated to examine interhospital variation in HAI. RESULTS Of an estimated 444,165 patients, 8% had HAI. Patients with HAI were older, had a greater burden of chronic diseases, and more commonly underwent CABG/valve or multivalve operations (all P < .001). HAI was independently associated with mortality (odds ratio 4.02; 95% confidence interval [CI], 3.67 to 4.40), non-home discharge (odds ratio 3.48; 95% CI, 3.21 to 3.78), and a cost increase of $23,000 (95% CI, $20,900 to $25,200). At 90 days, HAI was associated with greater hazard of readmission (1.29; 95% CI, 1.24 to 1.35). Pulmonary infections had the greatest incremental impact on patient-level costs ($24,500; 95% CI, $23,100 to $26,000) and annual cohort costs ($121.8 million; 95% CI, $102.2 to $142.9 million). Significant hospital level variation in HAI was evident, with observed-to-expected ratios ranging from 0.17 to 4.30 for cases performed in 2018. CONCLUSIONS Infections after cardiac surgery remain common and are associated with inferior outcomes and increased resource use. Interhospital variation in this contemporary cohort emphasizes the ongoing need for systematic approaches in their prevention and management.

Timing of surgical revascularization for acute coronary syndrome remains debated. We assessed the impact of timing to coronary artery bypass grafting (CABG) on mortality and resource utilization in a national cohort.Adults admitted for acute coronary syndrome in the 2009-2018 National Inpatient Sample were grouped by time from coronary angiography to CABG (Δt): 0, 1-3, 4-7, and7 days. Generalized linear models were fit to evaluate associations between Δt and in-hospital mortality and hospitalization costs. Timing and mortality of CABG for acute coronary syndrome were compared between high-performing hospitals (below the median risk adjusted mortality for all CABG and valve operations) and others.Of 444,065 patients, Δt = 0 days in 12.3%, Δt = 1-3 days in 57.3%, Δt = 4-7 days in 26.3%, and Δt7 days in 4.2%. Risk-adjusted mortality was greatest at Δt = 0 days (4.5%, 95% confidence interval [CI], 4.1%-4.9%) and Δt7 days (4.0%, 95% CI 3.4%-4.7%), but similar for operations performed at Δt = 1-3 days (1.8%, 95% CI 1.7%-1.9%) and Δt = 4-7 days (2.1%, 95% CI 1.9%-2.3%). Compared to Δt = 1-3 days, hospitalization costs were greater by $6,400 (95% CI $5,900-$6,900) for Δt = 4-7 days and $21,200 (95% CI $19,800-$22,600) for Δt7 days. High-performing hospitals had similar time to CABG as others (2 vs 2 days, P = .17), but lower mortality (0.9% vs 3.3%, P.001).Revascularization on days 1-3 and 4-7 led to comparable in-hospital mortality, with greater rates on day 0 and after day 7. Costs were greater for revascularization at days 4-7 compared with days 1-3. These findings support the reduction of time to revascularization to 1-3 days when deemed clinically appropriate and feasible.

Despite minimum volume recommendations, the majority of esophagectomies are performed at centers with fewer than 20 annual cases. The present study examined the impact of institutional esophagectomy volume on in-hospital mortality, complications, and resource use after esophageal resection.The 2010-2018 Nationwide Readmissions Database was queried to identify all adult patients undergoing esophagectomy for malignancy. Hospitals were categorized as a high-volume hospital (HVH) if performing at least 20 esophagectomies annually and as a low-volume hospital (LVH) if performing fewer than 20 esophagectomies annually. Multivariable models were developed to study the impact of volume on outcomes of interest, which included in-hospital mortality, complications, duration of hospitalization, inflation adjusted costs, readmissions, and nonhome discharge.Of an estimated 23,176 hospitalizations, 45.6% occurred at HVHs. Incidence of esophagectomy increased significantly along with median institutional caseload over the study period, while the proportion on hospitals considered HVHs remained steady at approximately 7.4%. After adjusting for relevant patient and hospital characteristics, HVH status was associated with decreased mortality (AOR, 0.65), length of stay (β = -1.83), pneumonia (AOR, 0.69), prolonged ventilation (AOR, 0.50), sepsis (AOR, 0.80), and tracheostomy (AOR, 0.66) but increased odds of nonhome discharge (AOR, 1.56; all P.01), with LVH status as reference.Many clinical outcomes of esophagectomy are improved with no increment in costs when performed at centers with an annual caseload of at least 20, as recommended by patient advocacy organizations. These findings suggest that centralization of esophageal resections to high-volume centers may be congruent with value-based care models.

Background Surgical re-exploration following cardiac surgery has been associated with increased in-hospital complications and mortality in limited series. The present study examined trends in reoperation and its impact on clinical outcomes and resource use in a nationally-representative cohort. We sought to determine patient and hospital factors associated with re-exploration and reoperative mortality, defined as failure-to-rescue-surgical (FTR-S). Methods Adult hospitalizations entailing cardiac operations (coronary artery bypass and/or valve) were identified using the 2005-2018 National Inpatient Sample. Procedures were tabulated using International Classification of Diseases codes. Hospitals were ranked into tertiles according to risk-adjusted mortality, with the lowest stratified as high-performing. Multivariable regression models examined factors associated with re-exploration as well as clinical outcomes including FTR-S and resource utilization. Results Of an estimated 3,490,245 hospitalizations, 78,003 (2.23%) required re-exploration with decreasing incidence over time. Valvular procedures, preoperative intra-aortic balloon pump and liver disease were associated with greater likelihood of re-exploration. Reoperation was associated with increased odds of mortality (adjusted odds ratio (AOR): 3.86, 95%CI: 3.61-4.12), perioperative complications and resource utilization. Increasing time from index operation to re-exploration was associated with higher odds of mortality (AOR:1.10/day, 95%CI: 1.07-1.12). High-performing hospitals were associated with lower odds of re-exploration (AOR: 0.88, 95%CI: 0.82-0.95) and FTR-S (AOR: 0.29, 95%CI: 0.23-0.35). Conclusions Surgical re-exploration following cardiac surgery has declined over time. High performing hospitals demonstrated lower rates of re-exploration and subsequent failure-to-rescue. Although unable to identify specific practices, our study highlights the presence of significant variation in takeback rates and further study of underlying factors is warranted.

Background Transsternal open thymectomy has long been the most widely used approach for thymectomy, but recent decades have seen the introduction of minimally invasive surgery (MIS), such as video-assisted thoracoscopic surgery (VATS) and robot-assisted thoracoscopic surgery (RATS) thymectomy. This retrospective cohort study provides a national comparison of trends, outcomes, and resource utilization of open, VATS, and RATS thymectomy. Methods Admissions for thymectomies from 2008 to 2014 were identified in the National Inpatient Sample. Patients were identified as undergoing open, VATS, or RATS thymectomy. Propensity score-matched analyses were used to compare overall complication rates, length of stay (LOS), and cost of VATS and RATS thymectomies. Results An estimated 23,087 patients underwent thymectomy during the study period: open in 16,025 (69%) and MIS in 7217 (31%). Of the MIS cohort, 4119 (18%) underwent VATS and 3097 (13%) underwent RATS. Performance of RATS and VATS thymectomy increased while that of open thymectomy declined. Baseline characteristics between VATS and RATS were similar, except more women underwent VATS thymectomy. No differences in LOS or overall complication rates were appreciable in this study. VATS was associated with the lowest cost of the 3 approaches. Conclusions Our findings demonstrate the increasing adoption of MIS and declining use of the open surgical approach for thymectomy. There are no differences in overall complication rates between RATS and VATS thymectomy, but RATS is associated with greater cost and lower cardiac complication rates.

Despite evidence supporting its early use in respiratory failure, tracheostomy is often delayed in cardiac surgical patients given concerns for sternal infection. This study assessed national trends in tracheostomy creation among cardiac patients and evaluated the impact of timing to tracheostomy on postoperative outcomes.We used the 2005 to 2015 National Inpatient Sample to identify adults undergoing coronary revascularization or valve operations and categorized them based on timing of tracheostomy: early tracheostomy (ET) (postoperative days 1-14) and delayed tracheostomy (DT) (postoperative days 15-30). Temporal trends in the timing of tracheostomy were analyzed, and multivariable models were created to compare outcomes.An estimated 33,765 patients (1.4%) required a tracheostomy after cardiac operations. Time to tracheostomy decreased from 14.8 days in 2005 to 13.9 days in 2015, sternal infections decreased from 10.2% to 2.9%, and in-hospital death also decreased from 23.3% to 15.9% over the study period (all P for trend.005). On univariate analysis, the ET cohort had a lower rate of sternal infection (5.2% vs 7.8%, P.001), in-hospital death (16.7% vs 22.9%, P.001), and length of stay (33.7 vs 43.6 days, P .001). On multivariable regression, DT remained an independent predictor of sternal infection (adjusted odds ratio, 1.35; P.05), in-hospital death (odds ratio, 1.36; P.001), and length of stay (9.1 days, P.001), with no difference in time from tracheostomy to discharge between the 2 cohorts (P = .40).In cardiac surgical patients, ET yielded similar postoperative outcomes, including sternal infection and in-hospital death. Our findings should reassure surgeons considering ET in poststernotomy patients with respiratory failure.

Deep venous thrombosis and pulmonary embolism are life-threatening complications after surgery, warranting prophylaxis. However prophylaxis is not uniformly practiced among cardiac surgical patients. This study aimed to characterize the national incidence, mortality, and costs associated with thromboembolism after cardiac surgery.The 2005 to 2015 National Inpatient Sample was used to identify all adult patients undergoing coronary artery bypass grafting or valve surgery. International Classification of Disease codes were used to identify patients with deep venous thrombosis and pulmonary embolism.Of approximately 3 million patients undergoing cardiac surgery, 1.62% developed deep venous thrombosis and 0.38% pulmonary embolism. Those with deep venous thrombosis and pulmonary embolism were more commonly women (33.2% and 36.2 vs 31.2%, P.001), older (68.1 and 66.0% vs 65.7 years, P.001), and had a higher Elixhauser comorbidity index (4.0 and 4.7 vs 3.7, P.001). Deep venous thrombosis and pulmonary embolism were associated with increased mortality (4.95% and 14.8% vs 2.67%, P.001). After adjustment for baseline differences, deep venous thrombosis was associated with an incremental increase in cost of $12,308, whereas pulmonary embolism was associated with $13,879 cost increase after cardiac surgery. Pulmonary embolism was an independent predictor of mortality (adjusted odds ratio, 3.39; 95% confidence interval, 2.74-4.18).The mortality and financial burden related to thromboembolism in cardiac surgery are significant. Prophylaxis may be indicated in cardiac surgery patients to improve quality of care and reduce healthcare costs. Future controlled randomized trials investigating the benefit of thromboembolism prophylaxis in cardiac surgery are warranted.

Because the rate of rehospitalization after major cardiac surgery has been reported up to 22%, an investigation of potential modifiable elements in the discharge process has led our group to evaluate whether the day of discharge affects readmission performance.Our institutional Society of Thoracic Surgeons registry was used to identify all adult patients undergoing elective cardiac operations from 2008 to 2016. Emergency, transplant, and mechanical assist patients were excluded. The primary outcome was all-cause readmission within 30 days of operation. Multivariable logistic regression was used to develop a risk-adjusted predictive model of readmission risk.Of 4,877 patients discharged from our institution, 20% were discharged on a weekend or holiday. The overall rehospitalization rate was 11.3%, with comparable readmission rates for weekday and weekend and holiday discharges (11.4 vs 10.9, p = 0.73). A greater proportion of patients are discharged to facilities on weekdays than on weekends and holidays (15.0% vs 5.7%, p0.001). Discharge to a facility is associated with a higher all-cause, unadjusted readmission rate (16.7% vs 12.7%, p = 0.01). After adjusting for patient comorbidities, operative performance, and postoperative complications, weekend or holiday discharge is not associated with worse readmission performance (adjusted odds ratio, 1.0; 95% confidence interval, 0.77 to 1.32).Cardiac surgical patients in the weekend and holiday discharge cohort did not have significantly higher odds of readmission regardless of operative type and discharge disposition. Allocation of resources to changing weekend staffing may be better allocated to surgical site infection prevention and outpatient intervention programs.