What drives variability in postoperative cardiac surgery transfusion rates?

OBJECTIVE: Wide interhospital variation exists in cardiac surgical postoperative transfusion rates. We aimed to compare transfusion rates at 2 hospitals and identify the institutional practice factors, unrelated to patient or operative characteristics, associated with postoperative transfusion rates. METHODS: Records for adult patients undergoing routine cardiac surgery at 2 hospitals (H and L) from February 2020 to August 2022 were analyzed. Patient and operative characteristics, preoperative and intensive care unit hemoglobin values, and postoperative transfusion rates were compared. Transfusion indication was recorded prospectively. Propensity matching was completed to assess comparability of patient populations. RESULTS: After propensity matching patients at H and L on age, procedure type, predicted morbidity or mortality, crossclamp time, preoperative hypertension, preoperative heart failure, and preoperative stroke, 2111 patients remained, with similar characteristics except hypertension. Matched results showed no significant differences in mortality, reoperation, or other major outcomes. Hospital H transfused 36% of patients (mean postoperative hemoglobin [Hb] 10.5) with 1483 units of packed red blood cells whereas hospital L transfused 12% of patients (mean postoperative Hb 9.4) with 198 units of packed red blood cells (P < .001). For all patients with a Hb >7.5, hospital H versus L transfused 27% versus 0.9% (P < .001). Hospital L's sole transfusion indication for pretransfusion hemoglobin trigger >7.5 was bleeding versus hospital H, which had multiple indications. When Hb concentration alone was the indication for transfusion, the threshold at hospital H was <7.5 g/dL versus <6 g/dL at hospital L. CONCLUSIONS: Variation in transfusion rates between hospitals H and L resulted from strict adherence at hospital L to a transfusion trigger of <6 g/dL with narrow indications for transfusions above that Hb concentration.

Trends in the Use of Recombinant Activated Factor VII and Prothrombin Complex Concentrate in Heart Transplant Patients in Virginia.

OBJECTIVES: To explore trends in intraoperative procoagulant factor concentrate use in patients undergoing heart transplantation (HTx) in Virginia. Secondarily, to evaluate their association with postoperative thrombosis. DESIGN: Patients who underwent HTx were identified using a statewide database. Trends in off-label recombinant activated factor VII (rFVIIa) use and on-label and off-label prothrombin complex concentrate (PCC) use were tested using the Mantel-Haenszel test. Multivariate logistic regression was used to test for an association between procoagulant factor concentrate administration and thrombosis. SETTING: Virginia hospitals performing HTx. PARTICIPANTS: Adults undergoing HTx between 2012 and 2022. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Among 899 patients who required HTx, 100 (11.1%) received off-label rFVIIa, 69 (7.7%) received on-label PCC, and 80 (8.9%) received off-label PCC. There was a downward trend in the use of rFVIIa over the 10-year period (p = 0.04). There was no trend in on-label PCC use (p = 0.12); however, there was an increase in off-label PCC use (p < 0.001). Patients who received rFVIIa were transfused more and had longer cardiopulmonary bypass time (p < 0.001). Receipt of rFVIIa was associated with increased thrombotic risk (odds ratio [OR] 1.92; 95% CI 1.12-3.29; p = 0.02), whereas on-label and off-label PCC use had no association with thrombosis (OR 0.98, 95% CI 0.49-1.96, p = 0.96 for on-label use; and OR 0.61, 95% CI 0.29-1.30, p = 0.20 for off-label use). CONCLUSIONS: Use of rFVIIa in HTx decreased over the past decade, whereas off-label PCC use increased. Receipt of rFVIIa was associated with thrombosis; however, patients who received rFVIIa were more severely ill, and risk adjustment may have been incomplete.

Statewide data on surgical ablation for atrial fibrillation: The data provide a path forward.

OBJECTIVE: Atrial fibrillation (AF), if left untreated, is associated with increased intermediate and long-term morbidity/mortality. Surgical treatment for AF is lacking standardization in patient selection and lesion set, despite clear support from multi-society guidelines. The aim of this study was to analyze a statewide cardiac surgery registry to establish whether or not there is an association between center volume and type of index procedure with performance of surgical ablation (SA) for AF, the lesion set chosen, and ablation technology used. METHODS: Adult, first-time, nonemergency patients with preoperative AF between 2014 and 2022 excluding standalone SA procedures from a statewide registry of Society of Thoracic Surgeons data were included (N = 4320). AF treatment variability by hospital volume (ordered from smallest to largest) and surgery type were examined with χ2 analyses. Hospital-level Spearman correlations compared hospital volume with proportion of AF patients treated with SA. RESULTS: Overall, 37% of patients with AF were ablated at the time of surgery (63% of mitral procedures, 26% of non-mitrals) and 15% had left atrial appendage management only. There was a significant temporal trend of increasing performance of SA for AF over time (Cochran-Armitage = 27.8; P < .001). Hospital cardiac surgery volume did not correlate with the proportion of AF patients treated with SA (rs = 0.19; P = .603) with a rate of SA below the state average for academic centers. Of cases with SA (n = 1582), only 43% had a biatrial lesion set. Procedures that involved mitral surgery were more likely to include a biatrial lesion set (χ2 = 392.3; P < .001) for both paroxysmal and persistent AF. Similarly, ablation technology use was variable by type of concomitant operation (χ2 = 219.0; P < .001) such that radiofrequency energy was more likely to be used in non-mitral procedures. CONCLUSIONS: These results indicate an increase in adoption of SA for AF over time. No association between greater hospital volume or academic status and performance of SA for AF was established. Similar to national data, the type of index procedure remains the most consistent factor in the decision to perform SA with a disconnect between AF pathophysiology and decision making on the type of SA performed. This analysis demonstrates a gap between evidence-based guidelines and real-world practice, highlighting an opportunity to confer the benefits of concomitant SA to more patients.

Unexpected impact of preoperative anemia in low-risk isolated coronary artery bypass grafting or single-valve surgical patients: Do not overlook these patients in anemia management!

OBJECTIVE: Preoperative anemia is prevalent in cardiac surgery and independently associated with increased risk for short-term and long-term mortality. The purpose of this study was to examine the effect of preoperative hematocrit (Hct) on outcomes in cardiac surgical patients and whether the effect is comparable across levels of Society of Thoracic Surgeons Predicted Risk of Mortality (STS PROM). METHODS: The study consisted of adult, isolated coronary artery bypass grafting (CABG) or single-valve surgical patients in a statewide registry from 2011 to 2022 (N = 29,828). Regressions were used to assess effect of preoperative Hct on STS-defined major morbidity/mortality including the interaction of Hct and STS PROM as continuous variables. RESULTS: Median age was 66 years (58-73 years), STS PROM was 1.02% (0.58%-1.99%), and preoperative Hct was 39.5% (35.8%-42.8%). The sample consisted of 78% isolated CABG (n = 23,261), 10% isolated mitral valve repair/replacement (n = 3119), 12% isolated aortic valve replacement (n = 3448), and 29% were female (n = 8646). Multivariable analyses found that greater Hct was associated with reduced risk of STS-defined morbidity/mortality (odds ratio, 0.96; P < .001). These effects for Hct persisted even after adjustment for intraoperative blood transfusion. The interaction of Hct and STS PROM was significant for morbidity/mortality (odds ratio, 1.01; P < .001). There was a stronger association between Hct levels and morbidity/mortality risk in the patients with the lowest STS risk compared with patients with the greatest STS risk. CONCLUSIONS: Patients with lower risk had a greater association between preoperative Hct and major morbidity and mortality compared with patients with greater risk. Preoperative anemia management is essential across all risk groups for improved outcomes.

Institutional Variability in Red Blood Cell Transfusion With Coronary Bypass in a Statewide Database.

BACKGROUND: The decision to perform transfusion is common but varies among centers and surgeons. This study looked at variables associated with red blood cell (RBC) transfusion in a statewide database. The study aimed to understand discrepancies in transfusion rates among hospitals and to establish whether the hospital itself was a significant variable in transfusion, independent of variables known to affect transfusion in patients undergoing cardiac surgical procedures. METHODS: The Maryland Cardiac Surgery Quality Initiative is a consortium of centers in the state. Patients undergoing isolated coronary artery bypass grafting from January 2018 to June 2020 from 10 centers in Maryland were included. Multivariable logistic regression was used to determine probability of RBC transfusion with covariates, including age, preoperative hemoglobin value, The Society of Thoracic Surgeons predicted risk of mortality, emergency status, preoperative adenosine diphosphate receptor blocker use, sex, body mass index, and off-pump status. RESULTS: A total of 5343 patients were included and had an overall RBC transfusion rate of 30.3% (range, 11.3%-55.8%). There was significant variability in the incidence of RBC transfusion among hospitals (χ2 = 604.7; P < .001). After covariate adjustment, a significant effect of hospital on transfusion remained (Wald = 547.3; P < .001). Hospital variation in RBC transfusion was not correlated with hospital variation in median age (P = .467), hemoglobin (P 0 855), The Society of Thoracic Surgeons predicted risk of mortality (P = .855), or sex (P = .726). CONCLUSIONS: In a statewide analysis, wide variability in transfusion rates was observed, with hospital-specific management strongly associated with RBC transfusion. This study suggests that RBC transfusion may be affected by the culture and practices of an institution independent of clinical and demographic variables.

Multiple Arterial Graft Use in Coronary Artery Bypass Surgery: Surgeon Perspective vs Practice.

BACKGROUND: Despite supportive evidence and guidelines, the use of multiple arterial grafts (MAGs) in coronary artery bypass grafting remains low. We sought to determine surgeon perception of personal MAG use and compare this with actual MAG use. METHODS: We conducted a statewide surgeon survey of MAG use, presence of a hospital MAG protocol, and barriers for MAG use, with a response rate of 78% (n = 25). Surgeon survey responses were compared with actual Society of Thoracic Surgeons patient data from January 1, 2017, to December 31, 2020 using χ2 or Fisher's exact tests. RESULTS: Of 5299 patients who had first-time, nonemergent, isolated coronary artery bypass grafting (≥2 grafts) by responding surgeons, 16% received MAG (n = 825). MAG use in patients whose surgeons self-designated as "routine" MAG users was 21% vs 7% for "nonroutine" users. Surgeons with a hospital protocol for MAG use utilized MAG more often (18% vs 14%, P = .001). Surgeons who were unconvinced by the data on the benefits of MAGs used MAGs in 11% vs 22% in surgeons who were convinced. MAG use increased over time, particularly from before to after the survey (13.1% vs 30.5%, P < .001). CONCLUSIONS: Although MAG use increased over time, barriers to routine use remain. In surgeons who reported routine use, only 21% of their patients received MAGs. Hospital protocols, education, and increased awareness may reduce barriers to use and encourage evidence-based clinical practice.

Association of Radial Artery Access with Reduced Incidence of Acute Kidney Injury.

Objectives: To determine if radial artery (RA) access compared with femoral artery (FA) access for percutaneous coronary intervention (PCI) is associated with a lower incidence of acute kidney injury (AKI). Background: AKI results in substantial morbidity and cost following PCI. Prior studies comparing the occurrence of AKI associated with radial artery (RA) versus femoral artery (FA) access have mixed results. Methods: Using a large state-wide database, 14,077 patients (8,539 with RA and 5,538 patents with FA access) were retrospectively compared to assess the occurrence of AKI following PCI. To reduce selection bias and balance clinical data across the two groups, a novel machine learning method called a Generalized Boosted Model was conducted on the arterial access site generating a weighted propensity score for each variable. A logistic regression analysis was then performed on the occurrence of AKI following PCI using the weighted propensity scores from the Generalized Boosted Model. Results: As shown in other studies, multiple variables were associated with an increase in AKI after PCI. Only RA access (OR 0.82; 95% CI 0.74-0.91) and male gender (OR 0.80; 95% CI 0.72-0.89) were associated with a lower occurrence of AKI. Based on the calculated Mehran scores, patients were stratified into groups with an increasing risk of AKI. RA access was consistently found to have a lower risk of AKI compared with FA access across these groups of increasing risk. Conclusions: Compared with FA access, RA access is associated with an 18% lower rate of AKI following PCI. This effect was observed among different levels of risk for developing AKI. Although developed from a retrospective analysis, this study supports the use of RA access when technically possible in a diverse group of patients.

Impact of Preoperative Hematocrit, Body Mass Index, and Red Cell Mass on Allogeneic Blood Product Usage in Adult Cardiac Surgical Patients: Report From a Statewide Quality Initiative.

OBJECTIVE: The study aims were to evaluate current blood transfusion practice in cardiac surgical patients and to explore associations between preoperative anemia, body mass index (BMI), red blood cell (RBC) mass, and allogeneic transfusion. DESIGN: Multicenter retrospective study. SETTING: Academic and non-academic centers. PARTICIPANTS AND INTERVENTIONS: After Institutional Review Board approval, 26,499 patients who underwent coronary artery bypass grafting ± valve replacement/repair between 2011 and 2019 were included from the Maryland Cardiac Surgery Quality Initiative database. Patients were stratified into BMI categories (<25, 25 to <30, and ≥30 kg/m2), and a multivariable logistic regression model was fit to determine if preoperative hematocrit, BMI, and RBC mass were associated independently with allogeneic transfusion. RESULTS: Preoperative anemia was found in 55.4%, and any transfusion was administered to 49.3% of the entire cohort. Females and older patients had lower BMI and RBC mass. Increased RBC and cryoprecipitate transfusions occurred more frequently after surgery in the lower BMI group. After adjustments, increased transfusion was associated with a BMI <25 relative to a BMI ≥30 at an odds ratio (OR) of 1.26 (95% confidence interval [CI]: 1.08-1.39). For each 1% increase in preoperative hematocrit, transfusion was decreased by 9% (OR: 0.91; 95% CI: 0.90-0.92). For every 500 mL increase in RBC mass, there was a 43% reduction of transfusion (OR: 0.57; 95% CI: 0.55-0.58). CONCLUSIONS: Transfusion probability modeling based on calculated RBC mass eliminated sex differences in transfusion risk based on preoperative hematocrit, and may better delineate which patients may benefit from more rigorous perioperative blood conservation strategy.

Barriers to atrial fibrillation ablation during mitral valve surgery.

BACKGROUND: Nearly 40% of patients with atrial fibrillation (AF) undergoing mitral valve surgery do not receive concomitant ablation despite societal guidelines. We assessed barriers to implementation of this evidence-based practice through a survey of cardiac surgeons in 2 statewide quality collaboratives. METHODS: Adult cardiac surgeons across 2 statewide collaboratives were surveyed on their knowledge and practice regarding AF ablation. Questions concerning experience, clinical practice, case scenarios, and barriers to implementation were included. RESULTS: Among 66 respondents (66 of 135; 48.9%), the majority reported "very comfortable/frequently use" cryoablation (53 of 66; 80.3%) and radiofrequency (55 of 66; 83.3%). Only 12.1% (8/66) were not aware of the recommendations. Approximately one-half of the respondents reported learning AF ablation in fellowship (50.0%; 33 of 66) or attending courses (47.0%; 31 of 66). Responses to clinical scenarios demonstrated wide variability in practice patterns. One-half of the respondents reported no barriers; others cited increased cross-clamp time, excessive patient risk, and arrhythmia incidence as obstacles. Desired interventions included cardiology/electrophysiology support, protocols, pacemaker rate information, and education in the form of site visits, videos and proctors. CONCLUSIONS: Knowledge of evidence-based recommendations and practice patterns vary widely. These data identify several barriers to implementation of concomitant AF ablation and suggest specific interventions (mentorship/support, protocols, research, and education) to overcome these barriers.

Hospital variability in modifiable factors driving coronary artery bypass charges.

OBJECTIVE: Coronary artery bypass grafting is associated with significant interhospital variability in charges. Drivers of hospital charge variability remain elusive. We identified modifiable factors associated with statewide interhospital variability in hospital charges for coronary artery bypass grafting. METHODS: Charge data were used as a surrogate for cost. Society of Thoracic Surgeons data from Maryland institutions and charge data from the Maryland Health Care Commission were linked to characterize interhospital charge variability for coronary artery bypass grafting. Multivariable linear regression was used to identify perioperative factors independently related to coronary artery bypass grafting charges. Of the factors independently associated with charges, we analyzed which factors varied between hospitals. RESULTS: A total of 10,337 patients underwent isolated coronary artery bypass grafting at 9 Maryland hospitals from 2012 to 2016, of whom 7532 patients were available for analyses. Mean normalized charges for isolated coronary artery bypass grafting varied significantly among hospitals, ranging from $30,000 to $57,000 (P < .001). Longer preoperative length of stay, operating room time, and major postoperative morbidity including stroke, renal failure, prolonged ventilation, reoperation, and deep sternal wound infection were associated with greater hospital charges. Incidence of major postoperative events, except stroke and deep sternal wound infection, was variable between hospitals. In a univariate linear regression model, patient risk profile only accounted for approximately 10% of statistical variance in charges. CONCLUSIONS: There is significant charge variability for coronary artery bypass grafting among hospitals within the same state. By targeting variation in preoperative length of stay, operating room time, postoperative renal failure, prolonged ventilation, and reoperation, cardiac surgery programs can realize cost savings while improving quality of care for this resource-intense patient population.

Isolated aortic valve replacement with bio-prostheses in patients age 50 to 65 years: a decade of statewide data on cost and patient outcomes.

BACKGROUND: Guidelines for choice of replacement valve-mechanical versus bio-prosthetic, are well established for patients aged <50 and >65 years. We studied the trends and implications of aortic valve replacement (AVR) with mechanical versus bioprosthetic valve in patients aged 50 to 65 years. METHODS: STS and cost database of 17 centers for isolated AVR surgery were analyzed by dividing them into bioprosthetic valve (BV) or mechanical valve (MV) groups. RESULTS: From 2002 to 2011, 3,690 patients had AVR, 18.6% with MV and 81.4% with BV. Use of BV for all ages increased from 71.5% in 2002 to 87% in 2011. There were 1127 (30.5%) patients in the age group 50-65 years. Use of BV in this group almost doubled, 39.6% in 2002 to 76.8% in 2011. Mean age of patients in BV group was higher (59.2±4.2 years vs. 56.7±4.3 years, P≤0.0001). Preoperative renal failure, heart failure and chronic obstructive pulmonary disease favored use of BV, whereas preoperative atrial fibrillation favored AVR with MV. Mortality (MV 2.2% vs. BV 2.36%) and other postoperative outcomes between the groups were similar. Cost of valve replacement increased for both groups (MV $26,191 in 2002 to $42,592 in 2011; BV $27,404 in 2002 to $44,257 in 2011). CONCLUSIONS: Use of bioprostheses for AVR has increased; this change is more pronounced in patients aged 50-65 years. Specific preoperative risk factors influence the choice of valve for AVR. Postoperative outcomes between the two groups were similar. Long-term implications of this changing practice, in particular, reoperation for bioprosthetic valve degeneration should be examined.

Number and Type of Blood Products Are Negatively Associated With Outcomes After Cardiac Surgery.

BACKGROUND: The association between blood transfusion and adverse outcome is documented in cardiac surgery. However, the incremental significance of each unit transfused, whether red blood cell (RBC) or non-RBC, is uncertain. This study examined the relationship of patient outcomes with the type and number of blood product units transfused. METHODS: Statewide data from 24 082 adult cardiac surgery patients were included. The relationship with blood transfusion was assessed for morbidity and 30-day mortality using total number of RBC and non-RBC units transfused, specific type of non-RBC units, and different combinations of transfusion (only RBC, only non-RBC, RBC + non-RBC). Multivariable logistic regressions examined these associations. RESULTS: Median age was 66 years (30% female patients), and 51% of patients received a transfusion (31%-66% across hospitals). Risk-adjusted analyses found each blood product unit was associated with 9%, 7%, and 4% greater odds for 30-day mortality, major morbidity, and minor morbidity, respectively (all P < .001). Odds for 30-day mortality were 13% greater with each RBC unit (P < .001) and 6% greater for each non-RBC unit (P < .001). Each unit of fresh frozen plasma (P < .001) and platelets (P < .001) increased the odds for 30-day mortality, but no effect was found for cryoprecipitate (P = .725). Odds for 30-day mortality were lower for non-RBC-only (odds ratio, 0.52; P = .030) and greater for RBC + non-RBC (odds ratio, 2.98; P < .001) compared with RBC-only transfusion. CONCLUSIONS: Independent of center variability on transfusion methods, each additional unit transfused was associated with increased odds for complications, with RBC transfusion carrying greater risk compared with non-RBC. Comprehensive evidence-based clinical approaches and coordination are needed to guide each blood transfusion event after cardiac surgery.

Access to Left Ventricular Assist Device: Travel Time Does Not Tell The Whole Story.

BACKGROUND: Negative health effects of traveling longer distances for surgical services have been reported. Given the high complexity of multidisciplinary care required for management of Left Ventricular Assist Device (LVAD) implantation, only 4 of 18 centers in our state perform these operations. Given the limited access we hypothesized increased travel time would adversely affect postoperative outcomes and 30-d mortality. METHODS: A statewide Society of Thoracic Surgeons database was queried to identify patients undergoing Heartmate II/III and HVAD implantation, and 725 patients were identified. Travel time was calculated by zip code. Patients were stratified into regional and distant groups by the upper quartile of travel time (1-h). Preoperative variables and outcomes were compared between the groups. Multivariate analysis was performed to evaluate the impact of travel time in risk-adjusted models of 30-d mortality. RESULTS: Median patient travel time to their LVAD center in our state is 32 min (mean 53 ± 65 min, 46 ± 71 miles). Patients in the distant group (n = 191) had lower median incomes, higher self-pay status, higher rates of medical comorbid disease. Despite these differences there was no difference between the groups in ICU and/or hospital length of stay, readmission, postoperative complications, or 30-d mortality. Multivariate regression demonstrated insurance status, age, and prior surgery predicted 30-d mortality, but not travel time. CONCLUSIONS: Despite only four centers in the state performing LVAD implantation, travel time was strongly associated with preoperative risk, and socioeconomic status but not postoperative outcomes or 30-d mortality. Therefore, increasing access should focus on insurance, and patient characteristics not travel time.

Clinical Practice Variation and Outcomes for Stanford Type A Aortic Dissection Repair Surgery in Maryland: Report from a Statewide Quality Initiative.

BACKGROUND: Stanford Type A aortic dissection repair surgery is associated with high mortality and clinical practice remains variable among hospitals. Few studies have examined statewide practice variation. METHODS: Patients who had Stanford Type A aortic dissection repair surgery in Maryland between July 1, 2014 and June 30, 2018 were identified using the Maryland Cardiac Surgery Quality Initiative (MCSQI) database. Patient demographics, comorbidities, surgery details, and outcomes were compared between hospitals. We also explored the impact of arterial cannulation site and brain protection technique on outcome. RESULTS: A total of 233 patients were included from eight hospitals during the study period. Seventy-six percent of surgeries were done in two high-volume hospitals (≥10 cases per year), while the remaining 24% were done in low-volume hospitals. Operative mortality was 12.0% and varied between 0 and 25.0% depending on the hospital. Variables that differed significantly between hospitals included patient age, the percentage of patients in shock, left ventricular ejection fraction, creatinine level, arterial cannulation site, brain protection technique, tobacco use, and intraoperative blood transfusion. The percentage of patients who underwent aortic valve repair or replacement procedures differed significantly between hospitals (p < 0.001), although the prevalence of moderate-to-severe aortic insufficiency was not significantly different (p = 0.14). There were no significant differences in clinical outcomes including mortality, renal failure, stroke, or gastrointestinal complications between hospitals or based on arterial cannulation site (all p > 0.05). Patients who had aortic cross-clamping or endovascualr repair had more embolic strokes when compared with patients who had hypothermic circulatory arrest (p = 0.03). CONCLUSION: There remains considerable practice variation in Stanford Type A aortic dissection repair surgery within Maryland including some modifiable factors such as intraoperative blood transfusion, arterial cannulation site, and brain protection technique. Continued efforts are needed within MCSQI and nationally to evaluate and employ the best practices for patients having acute aortic dissection repair surgery.

Racial Disparity in Cardiac Surgery Risk and Outcome: Report From a Statewide Quality Initiative.

BACKGROUND: Racial disparities persist in health care. Our study objective was to evaluate racial disparity in cardiac surgery in Maryland. METHODS: A statewide database was used to identify patients. Demographics, comorbidities, and predicted risk of death were compared between races. Crude mortality and incidence of complications were compared between groups, as were risk-adjusted odds for mortality and major morbidity or mortality. RESULTS: The study included 23,094 patients. Most patients were white (75.8%), followed by African American (16.3%), Asian (3.8%), and other races (4.1%). African Americans had a higher preoperative risk for mortality based on The Society of Thoracic Surgeons predictive models compared with white patients (3.0% vs 2.3%, P < .001). African Americans also had higher prevalence of diabetes mellitus, hypertension, peripheral vascular disease, and cerebral vascular disease than white patients. After adjustment for preoperative risk, there was no difference in 30-day mortality between African Americans (odds ratio [OR], 1.26; 95% confidence interval [CI], 0.99-1.59), Asians (OR, 1.22; 95% CI, 0.75-1.97), and other races (OR, 1.18; 95% CI, 0.74-1.89) compared with whites. African Americans had lower risk-adjusted odds of major morbidity or mortality compared with whites (OR, 0.83; 95% CI, 0.75-0.93). CONCLUSIONS: African American cardiac surgical patients have the highest preoperative risk in Maryland. Patients appeared to receive excellent cardiac surgical care, regardless of race, as risk-adjusted mortality did not differ between groups, and African American patients had lower risk-adjusted odds of major morbidity or mortality than white patients. Future interventions in Maryland should be aimed at reducing preoperative risk disparity in cardiac surgical patients.

Risk Aversion in Cardiac Surgery: 15-Year Trends in a Statewide Analysis.

BACKGROUND: With a rising emphasis on public reporting, we hypothesized that select hospitals are becoming increasingly risk-averse by avoiding high-risk operations. Further, we evaluated the association between risk-averse practices, outcomes, and publicly reported quality measures. METHODS: Clinical data from 78,417 patients undergoing cardiac surgery (2002-2016) from a regional consortium was paired with publicly available reimbursement and quality data. High-risk surgery was defined as predicted risk of mortality ≥5%. Hospital risk aversion was defined as a significant decrease in both high-risk volume and proportion, with cases stratified by hospital risk aversion status for univariate analysis. RESULTS: The rate of high-risk cases decreased from 17.9% in 2002 to 12.6% in 2016. Significant risk aversion was seen in 39% of hospitals, which had a 59% decrease in high-risk volume vs a 16% decrease at non-risk-averse hospitals. In the last 5 years, declining high-risk cases at risk-averse hospitals were driven by fewer cases from transfers (19.2% vs 28.1%, P < .001) and the emergency department (17.6% vs 19.2%, P = .001). Only non-risk-averse hospitals had mortality rates lower than expected (risk-averse: 0.97 [95% confidence interval, 0.91-1.03], P = .30; non-risk-averse: 0.88 [95% confidence interval, 0.83-0.94], P = .001). There were no differences by risk aversion status in reported ratings or financial incentives (all P > .05). CONCLUSIONS: Over 60% of hospitals continue to operate on high-risk patients, with concentration of care driven by transfer patterns. These non-risk-averse hospitals are high-performing with better-than-expected outcomes, particularly in high-risk cases. Transparency and objectivity in reporting are essential to ensure continued access for these high-risk patients.

Current quality reporting methods are not adequate for salvage cardiac operations.

OBJECTIVES: Outcomes in cardiac surgery are benchmarked against national Society of Thoracic Surgeons (STS) data and include patients undergoing elective, urgent, emergent, and salvage operations. This practice relies on accurate risk adjustment to avoid risk-averse behavior. We hypothesize that the STS risk calculator does not adequately characterize the risk of salvage operations because of their heterogeneity and infrequent occurrence. METHODS: Data on all cardiac surgery patients with an STS predicted risk score (2002-2017) were extracted from a regional database of 19 cardiac surgery centers. Patients were stratified according to operative status for univariate analysis. Observed-to-expected (O:E) ratios for mortality and composite morbidity/mortality were calculated and compared among elective, urgent, emergent, and salvage patients. RESULTS: A total of 76,498 patients met inclusion criteria. The O:E mortality ratios for elective, urgent, and emergent cases were 0.96, 0.98, and 0.93, respectively (all P values > .05). However, mortality rate was significantly higher than expected for salvage patients (O:E ratio, 1.41; P = .04). Composite morbidity/mortality rate was lower than expected in elective (O:E ratio, 0.81; P = .0001) and urgent (O:E ratio, 0.93; P = .0001) cases but higher for emergent (O:E ratio, 1.13; P = .0006) and salvage (O:E ratio, 1.24; P = .01). O:E ratios for salvage mortality were highly variable among each of the 19 centers. CONCLUSIONS: The current STS risk models do not adequately predict outcomes for salvage cardiac surgery patients. On the basis of these results, we recommend more detailed reporting of salvage outcomes to avoid risk aversion in these potentially life-saving operations.

Impact of transfer status on real-world outcomes in nonelective cardiac surgery.

OBJECTIVE: Transfer from hospital to hospital for cardiac surgery represents a large portion of some clinical practices. Previous literature in other surgical fields has shown worse outcomes for transferred patients. We hypothesized that transferred patients would be higher risk and demonstrate worse outcomes than those admitted through the emergency department. METHODS: All patients undergoing cardiac operations with a Society of Thoracic Surgeons Predicted Risk of Mortality were evaluated from a multicenter, statewide Society of Thoracic Surgeons database. Only patients requiring admission before surgery were included. Patients were stratified by admission through the emergency department or in transfer. Transfers were further stratified by the cardiothoracic surgery capabilities at the referring center. RESULTS: A total of 13,094 patients met the inclusion criteria of admission before surgery. This included 7582 (57.9%) transfers, of which 502 (6.6%) were referred from cardiac centers. Compared with emergency department admissions, transfers had increased hospital costs despite lower operative risk (Predicted Risk of Mortality 1.5% vs 1.6%, P < .01) and equivalent postoperative morbidity (15.6% vs 15.3% P = .63). In risk-adjusted analysis, transfer status was not independently associated with worse outcomes. Patients transferred from centers that perform cardiac surgery are higher risk than general transfers (Predicted Risk of Mortality 2.5% vs 1.5, P < .01), but specialized care results in excellent risk-adjusted outcomes (observed/expected: mortality 0.81; morbidity or mortality 0.90). CONCLUSIONS: Transfer patients have similar rates of postoperative complications but increased resource use compared with patients admitted through the emergency department. Patients transferred from centers that perform cardiac surgery represent a particularly high-risk subgroup.

Mitigating the Risk: Transfusion or Reoperation for Bleeding After Cardiac Surgery.

BACKGROUND: Several studies have established morbidity associated with bleeding after cardiac surgery. Although reoperation has been implicated as the marker for this morbidity, there remains limited understanding regarding relative morbidities of reoperation and substantial transfusion. METHODS: The Society of Thoracic Surgeons (STS) Maryland Adult Cardiac Surgery Database (July 2011-September 2018) was reviewed (N = 23,240). Substantial transfusion was defined as requiring greater than the reoperation group median red blood cells (5 units) and non-red blood cells (4 units). Patients were stratified into 4 subgroups: group 1, no reoperation without substantial transfusion (n = 22,365); group 2, reoperation without substantial transfusion (n = 351); group 3, no reoperation with substantial transfusion (n = 350); and group 4, reoperation with substantial transfusion (n = 167). Operative morbidity and mortality were compared. RESULTS: Reoperation patients were older with a higher STS predicted risk of mortality (1.8% vs 1.2%, P < .001). Multivariable analysis demonstrated that group 4 increased the odds of renal failure (odds ratio [OR] 7.36, P < .001), stroke (OR 3.24, P = .002), and operative mortality (OR 8.68, P < .001) compared with group 1. Both group 2 and group 3 increased the odds of mortality and renal failure compared with group 1. However, group 3 had greater risk for renal failure (OR 3.48, P < .001) and mortality (OR 2.91, P < .001) than group 2. CONCLUSIONS: Although reoperation for bleeding is associated with morbidity after cardiac surgery, substantial transfusion without reoperation appears to increase morbidity compared with a limited-transfusion reoperative approach. Better timing for reoperation and guided transfusion approaches may mitigate morbidity compared with substantial transfusion alone.

Primary graft dysfunction after heart transplantation: Outcomes and resource utilization.

BACKGROUND: A unified definition of primary graft dysfunction (PGD) after heart transplantation was adopted in 2014, with moderate and severe PGD defined as a need for mechanical circulatory support. While risk factors for PGD are well identified, outcomes and resource utilization have not been well-studied. We examined the resource utilization and associated costs with PGD. METHODS: All adult heart transplantations (2001-2016) from a statewide Society of Thoracic Surgery database were analyzed by dividing them into two groups-with PGD (requiring mechanical circulatory support) and without PGD. RESULTS: Of the 718 heart transplants, 110 (15.3%) patients developed PGD. Prevalence of PGD for the study duration ranged from 3.7% to 22.7% with no significant trend. The most frequently used mechanical circulatory support device was intra-aortic balloon pump (88%), followed by extracorporeal membrane oxygenation (17%), and catheter-based circulatory support devices (3%). There were no significant differences in demographics or preoperative variables between the two groups. Resource utilization such as total intensive care unit hours, ventilation hours, reoperation for bleeding, blood product transfusions, and length of stay were significantly higher in the PGD group. Postoperative complications were also higher in PGD group including operative mortality (31.8% vs 3.8%, P < .0001). The median cost of heart transplantation was significantly higher in the PGD group $229 482 ($126 044-$388 889) vs $101 788 ($72 638-$181 180) P < .0001. CONCLUSION: Primary graft dysfunction following heart transplantation developed in 15% of patients. Patients with PGD had significantly higher complications, resource utilization, and mortality. Preventive measures to address the development of PGD would reduce resource utilization and improve outcomes.