Why Are We Weighting? Understanding the Estimates From Propensity Score Weighting and Matching Methods.

BACKGROUND: Propensity score methods are used in observational studies to compensate for the lack of random allocation by balancing measured baseline characteristics between treated and untreated patients. We sought to explain the treatment effect estimates derived from different propensity score methods. METHODS: We performed a retrospective analysis of long-term mortality after single internal mammary artery versus bilateral internal mammary artery (BIMA) conduit in 47 984 index isolated coronary artery bypass grafting procedures from 1992 to 2014 in the Northern New England Cardiovascular Disease Study Group registry using multivariable Cox regression, 1:1 propensity score matching, inverse probability weighting (IPW) among the treated, and IPW among the overall population treatment estimates. RESULTS: The mean duration of follow-up was 13.2 (interquartile range, 7.4-17.7) years. In multivariable Cox regression, the adjusted hazard ratio for mortality was 0.83 (95% CI, 0.75-0.92) in patients receiving BIMA compared with a single internal mammary artery. The 1:1 propensity matched (hazard ratio, 0.79 [95% CI, 0.69-0.91]) and IPW among the treated (hazard ratio, 0.83 [95% CI, 0.75-0.92]) estimates showed a protective treatment effect of BIMA use on mortality. However, the IPW estimate of treatment effect for the overall population showed an increased risk of mortality after BIMA that was not statistically significant (hazard ratio, 1.08 [95% CI, 0.94-1.24]). CONCLUSIONS: While the multivariable Cox regression, 1:1 propensity matching, and IPW treatment effect in the treated estimates demonstrate that BIMA was associated with a statistically significantly decreased risk of mortality, the IPW treatment effect in the average study population showed an increased risk of mortality associated with BIMA that was not statistically significant. This is attributed to the different populations (weighted to look like the overall study population versus treated group) represented by the 2 IPW approaches. Determining how the study population is balanced is a large driver of the treatment effect. Ultimately, the treatment effect estimate desired should drive the choice of the propensity score method.

A multi-center analysis of readmission after cardiac surgery: Experience of The Northern New England Cardiovascular Disease Study Group.

BACKGROUND: Readmissions after cardiac surgery are common and associated with increased morbidity, mortality and cost of care. Policymakers have targeted coronary artery bypass grafting to achieve value-oriented health care milestones. We explored the causes of readmission following cardiac surgery among a regional consortium of hospitals. METHODS: Using administrative data, we identified patients readmitted to the same institution within 30 days of cardiac surgery. We performed standardized review of readmitted patients' medical records to identify primary and secondary causes of readmission. We evaluated causes of readmission by procedure and tested for univariate associations between characteristics of readmitted patients and nonreadmitted patients in our clinical registry. RESULTS: Of 2218 cardiac surgery patients, 272 were readmitted to the index hospital within 30 days for a readmission rate of 12.3%. Median time to readmission was 9 days (interquartile range 4-16 days) and only 13% of patients were evaluated in-office before readmission. Readmitted patients were more likely to have had valve surgery (31.3% vs 22.7%) than patients not readmitted. Readmitted patients were also more likely to have preoperative creatinine more than or equal to 2 mg/dL (P = .015) or congestive heart failure (CHF) (P = .034), require multiple blood transfusions or sustained inotropic support (P < .001), and experience postoperative atrial fibrillation (P = .022) or renal insufficiency (P < .001). Infection (26%), pleural or pericardial effusion (19%), arrhythmia (16%), and CHF (11%) were the most common primary etiologies leading to readmission. CONCLUSIONS: Ensuring early follow-up for high-risk patient groups while improving early detection and management of the principal drivers of readmission represent promising targets for decreasing readmission rates.

Intensity of Glycemic Control Affects Long-Term Survival After Coronary Artery Bypass Graft Surgery.

BACKGROUND: A patient's hemoglobin (Hb) A1c level, regardless of diabetic status, is a measure of glycemic control. Studies have found it is an independent predictor of short-term death in patients undergoing coronary artery bypass grafting (CABG). In this study, we used preoperative HbA1c to assess whether levels are associated with short-term and long-term survival after CABG. METHODS: From a regional registry of consecutive cases, we identified 6,415 patients undergoing on-pump isolated CABG from 2008 to 2015 with documented preoperative HbA1c level. We defined four HbA1c groups: less than 5.7% (n = 1,713), 5.7% to 6.4% (n = 2,505), 6.5% to 8.0% (n = 1,377), and more than 8% (n = 820). Relationship to in-hospital outcomes and long-term survival was assessed. Outcome rates and hazard ratios were adjusted for patient and disease risk factors using multivariable logistic regression and Cox models. RESULTS: The study included 3,740 patients (58%) not diagnosed as having diabetes and 2,674 with diabetes. Prediabetes (HbA1c 5.7% to 6.4%) was documented in 52% (n = 1,933) of nondiabetic patients. Higher HbA1c values were associated with younger age, female sex, greater body mass index, more comorbid diseases, lower ejection fraction, more 3-vessel coronary disease, and recent myocardial infarction (p < 0.05 trend for all). After adjustment for patient risk, greater HbA1c values were not associated with higher rates of in-hospital death or morbidity. Long-term survival was significantly worse as HbA1c increased. Risk of death increased by 13% for every unit increase in HbA1c (adjusted hazard ratio, 1.13; 95% confidence interval, 1.07 to 1.19; p < 0.001). CONCLUSIONS: Preadmission glycemic control, as assessed by HbA1c, is predictive of long-term survival, with higher levels associated with poorer prognosis. Whether this risk can be modified by better glycemic control postoperatively remains to be determined.

Surgical Atrial Fibrillation Ablation Improves Long-Term Survival: A Multicenter Analysis.

BACKGROUND: The Society of Thoracic Surgeons guidelines recommend surgical ablation (SA) at the time of concomitant mitral operations, aortic valve replacement, coronary artery bypass grafting (CABG), and aortic valve replacement plus CABG for patients in atrial fibrillation (AF). The goal of this analysis was to assess the influence of SA on long-term survival. METHODS: A retrospective analysis of 20,407 consecutive CABG or valve procedures from 2008 to 2015 among seven centers reporting to a prospectively maintained clinical registry was conducted. Patients undergoing operation with documented preoperative AF were included (n = 2,740). Patients receiving SA were compared with patients receiving no SA. The primary end point was all-cause mortality. Secondary end points included in-hospital morbidity and mortality. RESULTS: The frequency of SA was 23.1% (n = 634), and an increase was seen in the rate of SA over the study period (p < 0.001). Concomitant SA was performed in 16.2% of CABG, 30.6% of valve, and 24.3% of valve plus CABG procedures. A substantial improvement was found in unadjusted survival among patients undergoing SA (hazard ratio 0.54, 95% confidence interval: 0.42 to 0.70). Moreover, no differences were found in postoperative complications. SA did have longer bypass times (p < 0.001) but a shorter overall length of stay (p < 0.001). After risk adjustment, SA patients had an improved 5-year survival (hazard ratio 0.69, 95% confidence interval: 0.51 to 0.92), and the effect was observed across all operations. CONCLUSIONS: In a multicenter cohort of patients with AF, concomitant SA resulted in substantially improved long-term survival across patients who underwent CABG, valve, and valve plus CABG. These findings support current guidelines from The Society of Thoracic Surgeons that recommend broader application of concomitant SA.

Comparative effectiveness of coronary artery bypass grafting versus percutaneous coronary intervention in a real-world Surgical Treatment for Ischemic Heart Failure trial population.

OBJECTIVE: There are no prospective randomized trial data to guide decisions on optimal revascularization strategies for patients with multivessel coronary artery disease and reduced ejection fraction. In this analysis, we describe the comparative effectiveness of coronary artery bypass grafting (CABG) versus percutaneous coronary intervention (PCI) in this patient population. METHODS: A multicenter, retrospective analysis of all CABG (n = 18,292) and PCIs (n = 55,438) performed from 2004 to 2014 among 7 medical centers reporting to the Northern New England Cardiovascular Disease Study Group. After applying inclusion and exclusion criteria from the Surgical Treatment for Ischemic Heart Failure trial, there were 955 CABG and 718 PCI patients with an ejection fraction ≤ 35% and 2- or 3-vessel disease. Inverse probability weighting was used for risk adjustment. The primary end point was all-cause mortality. Secondary end points included rates of 30-day mortality, stroke, acute kidney injury, and incidence of repeat revascularization. RESULTS: The median duration of follow-up was 4.3 years (range, 1.59-6.71 years). CABG was associated with improved long-term survival compared with PCI after risk adjustment (hazard ratio, 0.59; 95% confidence interval, 0.50-0.71; P < .01). Although CABG and PCI had similar 30-day mortality rates (P = .14), CABG was associated with a higher frequency of stroke (P < .001) and acute kidney injury (P < .001), whereas PCI was associated with a higher incidence of repeat revascularization (P < .001). CONCLUSIONS: Among patients with reduced ejection fraction and multivessel disease, CABG was associated with improved long-term survival compared with PCI. CABG should be strongly considered in patients with ischemic cardiomyopathy and multivessel coronary disease.

Coronary Revascularization With Single Versus Bilateral Mammary Arteries: Is It Time to Change?

BACKGROUND: Arterial conduits are preferred to venous conduits for coronary artery bypass grafting because of longer patency. A single internal mammary artery (SIMA) is used routinely. Bilateral internal mammary arteries (BIMA) are used less frequently. We sought to determine whether BIMA were superior to SIMA. METHODS: From our regional registry of consecutive open heart operations, we identified 47,984 patients who underwent isolated coronary artery bypass grafting from 1992 to 2014. Of the 1,482 BIMA patients, 1,297 were propensity matched to a cohort of SIMA patients. Short-term outcomes were compared using standard statistical techniques. Long-term survival was compared using Kaplan-Meier estimators and compared using a log-rank test. RESULTS: BIMA patients were younger and had fewer comorbid conditions than SIMA patients. After propensity weighting, BIMA and SIMA patients were well matched. There was no difference in in-hospital outcomes for BIMA versus SIMA patients for mortality (1.2% [n = 15] vs 0.8% [n = 10], p = 0.315), stroke (0.7% [n = 9] vs 0.7% [n = 9), p = 1.000), bleeding (2.2% [n = 28] vs 2.8% [n = 36], p = 0.311), or mediastinitis (0.8% [n = 10] vs 0.9% [n = 12], p = 0.667). The median follow-up was 12 years. Survival was better for BIMA than SIMA (adjusted hazard ratio, 0.79; 95% confidence interval, 0.69 to 0.91; p < 0.001). Survival curves began to separate after 5 years. At 15 years, the absolute difference in survival was 8.4%. CONCLUSIONS: In a large regional experience, BIMA is associated with no upfront risk of adverse events and improved long-term survival compared with SIMA. Our results indicate that BIMA conduits should be considered more frequently during coronary artery bypass grafting due to their demonstrated survival advantage.

Should Diabetes Be a Contraindication to Bilateral Internal Mammary Artery Grafting?

BACKGROUND: This study evaluates the influence of bilateral internal mammary artery (BIMA) versus single internal mammary artery (SIMA) grafting on postoperative morbidity and long-term survival among diabetic patients undergoing coronary artery bypass grafting (CABG). METHODS: A multicenter, retrospective analysis of 47,984 consecutive CABGs performed from 1992 to 2014 at 7 medical centers was conducted. Among the study population, 1,482 CABGs with BIMA were identified, and 1,297 BIMA patients were propensity-matched to 1,297 SIMA patients. The study cohort for this analysis, drawn from matched data, included 430 diabetic patients: 217 SIMA and 213 BIMA. The primary endpoint was long-term survival. Secondary endpoints included postoperative morbidity, length of stay, and in-hospital mortality. RESULTS: The median duration of follow-up was 9.3 (range, 4.3 to 13.9) years. Among propensity-matched diabetic patients, there was no significant difference in age, body mass index, or major baseline comorbidities. The groups were also well matched on the number of diseased coronary arteries and number of distal anastomoses performed. There was no difference in the rate of mediastinitis or sternal dehiscence (p = 0.503) or in-hospital mortality (p = 0.758) between groups. Both groups had a similar median length of stay of 5 (range, 4 to 7) days. Diabetic patients who received a BIMA had significantly improved long-term survival when compared with SIMA patients (hazard ratio 0.75 [95% confidence interval 0.57 to 0.98], p = 0.034). CONCLUSIONS: Among diabetics undergoing CABG, use of BIMA grafting does not result in increased in-hospital morbidity or mortality and confers a long-term survival advantage when compared with SIMA grafting. Thus, diabetic patients should be considered for BIMA grafting more frequently.

Does Use of Bilateral Internal Mammary Artery Grafting Reduce Long-Term Risk of Repeat Coronary Revascularization? A Multicenter Analysis.

BACKGROUND: Although previous studies have demonstrated that patients receiving bilateral internal mammary artery (BIMA) conduits during coronary artery bypass grafting have better long-term survival than those receiving a single internal mammary artery (SIMA), data on risk of repeat revascularization are more limited. In this analysis, we compare the timing, frequency, and type of repeat coronary revascularization among patients receiving BIMA and SIMA. METHODS: We conducted a multicenter, retrospective analysis of 47 984 consecutive coronary artery bypass grafting surgeries performed from 1992 to 2014 among 7 medical centers reporting to a prospectively maintained clinical registry. Among the study population, 1482 coronary artery bypass grafting surgeries with BIMA were identified, and 1297 patients receiving BIMA were propensity-matched to 1297 patients receiving SIMA. The primary end point was freedom from repeat coronary revascularization. RESULTS: The median duration of follow-up was 13.2 (IQR, 7.4-17.7) years. Patients were well matched by age, body mass index, major comorbidities, and cardiac function. There was a higher freedom from repeat revascularization among patients receiving BIMA than among patients receiving SIMA (hazard ratio [HR], 0.78 [95% CI, 0.65-0.94]; P=0.009). Among the matched cohort, 19.4% (n=252) of patients receiving SIMA underwent repeat revascularization, whereas this frequency was 15.1% (n=196) among patients receiving BIMA (P=0.004). The majority of repeat revascularization procedures were percutaneous coronary interventions (94.2%), and this did not differ between groups (P=0.274). Groups also did not differ in the ratio of native versus graft vessel percutaneous coronary intervention (P=0.899), or regarding percutaneous coronary intervention target vessels; the most common targets in both groups were the right coronary (P=0.133) and circumflex arteries (P=0.093). In comparison with SIMA, BIMA grafting was associated with a reduction in all-cause mortality at 12 years of follow-up (HR, 0.79 [95% CI, 0.69-0.91]; P=0.001), and there was no difference in in-hospital morbidity. CONCLUSIONS: BIMA grafting was associated with a reduced risk of repeat revascularization and an improvement in long-term survival and should be considered more frequently during coronary artery bypass grafting.

Improving patients' readiness for coronary artery bypass graft surgery.

BACKGROUND: Preoperative interventions improve outcomes for patients after coronary artery bypass surgery (CABG). OBJECTIVE: To reduce mortality for patients undergoing urgent CABG. METHODS: Eight centers implemented preoperative aspirin and statin, preinduction heart rate less than 80/min, hematocrit greater than 30%, blood sugar less than 150 mg/dL (8.3 mmol/L), and delayed surgery at least 3 days after a myocardial infarction. Data were collected on the last 150 isolated, urgent CABGs at each center (n=1200). A "bundle" score of 0 to 100 was calculated for each patient to represent the percentage of interventions used. RESULTS: Scores ranged from 33 to 100. About 56% of patients had a perfect score. Crude mortality and composite rates were lower in patients with higher scores, but once adjusted for patient and disease characteristics, the difference in scores was not significant. Higher scores were associated with shorter intubation: 6.0 hours (score 100), 8.0 hours (score 80-99), 8.4 hours (score<80) (log-rank P<.001). Median length of stay was shorter for patients with higher scores: 5 days (score 100), 6 days (scores 80-99), and 6 days (scores <80) (log-rank P<.001). CONCLUSION: Implementation of interventions to optimize patients' "readiness for surgery" is associated with shorter intubation times and shorter hospital stays after CABG.

Geographic variability in potentially discretionary red blood cell transfusions after coronary artery bypass graft surgery.

OBJECTIVE: A number of established regional quality improvement collaboratives have partnered to assess and improve care across their regions under the umbrella of the Cardiac Surgery Quality Improvement (IMPROVE) Network. The first effort of the IMPROVE Network has been to assess regional differences in potentially discretionary transfusions (<3 units red blood cells [RBCs]). METHODS: We examined 11,200 patients undergoing isolated nonemergent coronary artery bypass graft surgery across 56 medical centers in 4 IMPROVE Network regions between January 2008 and June 2012. Each center submitted the most recent 200 patients who received 0, 1, or 2 units of RBC transfusion during the index admission. Patient and disease characteristics, intraoperative practices, and percentage of patients receiving RBC transfusions were collected. Region-specific transfusion rates were calculated after adjusting for pre- and intraoperative factors among region-specific centers. RESULTS: There were small but significant differences in patient case mix across regions. RBC transfusions of 1 or 2 units occurred among 25.2% of coronary artery bypass graft procedures (2826 out of 11,200). Significant variation in the number of RBC units used existed across regions (no units, 74.8% [min-max, 70.0%-84.1%], 1 unit, 9.7% [min-max, 5.1%-11.8%], 2 units, 15.5% [min-max, 9.1%-18.2%]; P < .001). Variation in overall transfusion rates remained after adjustment (9.1%-31.7%; P < .001). CONCLUSIONS: Delivery of small volumes of RBC transfusions was common, yet varied across geographic regions. These data suggest that differences in regional practice environments, including transfusion triggers and anemia management, may contribute to variability in RBC transfusion rates.

Impact of preoperative left ventricular ejection fraction on long-term survival after aortic valve replacement for aortic stenosis.

BACKGROUND: The survival of patients who undergo aortic valve replacement (AVR) for severe aortic stenosis with reduced preoperative ejection fractions (EFs) is not well described in the literature. METHODS AND RESULTS: Patients undergoing AVR for severe aortic stenosis were analyzed using the Northern New England Cardiovascular Disease Study Group surgical registry. Patients were stratified by preoperative EF (≥50%, 40%-49%, and <40%) and concomitant coronary artery bypass grafting. Crude and adjusted survival across strata of EF was estimated for patients up to 8 years beyond their index admission. A total of 5277 patients underwent AVR for severe aortic stenosis between 1992 and 2008. There were 727 (14%) patients with preoperative EF <40%. Preoperative EF had minimal effect on postoperative morbidity. There was no difference in 30-day mortality across EF strata among the isolated AVR cohort. Preserved EF conferred 30-day survival benefit among the AVR+coronary artery bypass grafting population (EF≥50%, 96%; EF<40%, 91%; P=0.003). Patients with preserved EF had significantly improved 6-month and 8-year survival compared with their reduced EF counterparts. CONCLUSIONS: Survival after AVR or AVR+coronary artery bypass grafting was most favorable among patients with preoperative preserved EF. However, patients with mild to moderately depressed EF experienced a substantial survival benefit compared with the natural history of medically treated patients. Furthermore, minor reductions of EF carried equivalent increased risk to those with more compromised function suggesting patients are best served when an AVR is performed before even minor reductions in myocardial function.

Variability in surgeons' perioperative practices may influence the incidence of low-output failure after coronary artery bypass grafting surgery.

BACKGROUND: Postoperative low-output failure (LOF) is an important contributor to morbidity and mortality after coronary artery bypass grafting surgery. We sought to understand which pre- and intra-operative factors contribute to postoperative LOF and to what degree the surgeon may influence rates of LOF. METHODS AND RESULTS: We identified 11 838 patients undergoing nonemergent, isolated coronary artery bypass grafting surgery using cardiopulmonary bypass by 32 surgeons at 8 centers in northern New England from 2001 to 2009. Our cohort included patients with preoperative ejection fractions >40%. Patients with preoperative intraaortic balloon pumps were excluded. LOF was defined as the need for ≥2 inotropes at 48 hours, an intra- or post-operative intraaortic balloon pumps, or return to cardiopulmonary bypass (for hemodynamic reasons). Case volume varied across the 32 surgeons (limits, 80-766; median, 344). The overall rate of LOF was 4.3% (return to cardiopulmonary bypass, 2.6%; intraaortic balloon pumps, 1.0%; inotrope usage, 0.8%; combination, 1.0%). The predicted risk of LOF did not differ across surgeons, P=0.79, and the observed rates varied from 1.1% to 10.2%, P<0.001. Patients operated by low-rate surgeons had shorter clamp and bypass times, antegrade cardioplegia, longer maximum intervals between cardioplegia doses, lower cardioplegia volume per anastomosis or minute of ischemic time, and less hot-shot use. Patients operated on by higher LOF surgeons had higher rates of postoperative acute kidney injury. CONCLUSIONS: Rates of LOF significantly varied across surgeons and could not be explained solely by patient case mix, suggesting that variability in perioperative practices influences risk of LOF.

Long-term outcomes of endoscopic vein harvesting after coronary artery bypass grafting.

BACKGROUND: Use of endoscopic saphenous vein harvesting has developed into a routine surgical approach at many cardiothoracic surgical centers. The association between this technique and long-term morbidity and mortality has recently been called into question. The present report describes the use of open versus endoscopic vein harvesting and risk of mortality and repeat revascularization in northern New England during a time period (2001 to 2004) in which both techniques were being performed. METHODS AND RESULTS: From 2001 to 2004, 8542 patients underwent isolated coronary artery bypass grafting procedures, 52.5% with endoscopic vein harvesting. Surgical discretion dictated the vein harvest approach. The main outcomes were death and repeat revascularization (percutaneous coronary intervention or coronary artery bypass grafting) within 4 years of the index admission. The use of endoscopic vein harvesting increased from 34% in 2001 to 75% in 2004. In general, patients undergoing endoscopic vein harvesting had greater disease burden. Endoscopic vein harvesting was associated with an increased adjusted risk of bleeding requiring a return to the operating room (2.4 versus 1.7; P=0.03) but a decreased risk of leg wound infections (0.2 versus 1.1; P<0.001). Use of endoscopic vein harvesting was associated with a significant reduction in long-term mortality (adjusted hazard ratio, 0.74; 95% confidence interval, 0.60 to 0.92) but a nonsignificant increased risk of repeat revascularization (adjusted hazard ratio, 1.29; 95% confidence interval, 0.96 to 1.74). Similar results were obtained in propensity-stratified analysis. CONCLUSIONS: During 2001 to 2004 in northern New England, the use of endoscopic vein harvesting was not associated with harm. There was a nonsignificant increase in repeat revascularization, and survival was not decreased.

Long-term survival of the very elderly undergoing aortic valve surgery.

BACKGROUND: Increasing numbers of the very elderly are undergoing aortic valve procedures. We describe the short- and long-term survivorship for this cohort. METHODS AND RESULTS: We conducted a cohort study of 7584 consecutive patients undergoing open aortic valve surgery without (51.1%; AVR) or with (48.9%; AVR + CABG) concomitant coronary artery bypass graft surgery between November 10, 1987 through June 30, 2006. Patient records were linked to the Social Security Administration's Death Master File. Survivorship was stratified by age and concomitant CABG surgery. During 39 835 person-years of follow-up, there were 2877 deaths. Among AVR, there were 3304 patients <80 years of age, 419 patients 80 to 84 years, and 156 patients > or =85 years (24 patients >90 years). Among AVR+CABG patients, there were 2890 patients <80 years of age, 577 patients 80 to 84 years, and 238 patients > or =85 years (22 patients >90 years). Median survivorship for patients undergoing isolated AVR was 11.5 years (<80 years), 6.8 years (80 to 84 years), 6.2 years (> or =85 years); for patients undergoing AVR+CABG, median survivorship was 9.4 years (<80 years), 6.8 years (80 to 84 years), and 7.1 years (> or =85 years). Among both procedures, adjusted survivorship was significantly different across strata of age (P<0.001). These findings are similar to life expectancy of the general population from actuarial tables: 80 to 84 years (7 years) and > or =85 years (5 years). CONCLUSIONS: Survivorship among octogenarians is favorable, with more than half the patients surviving more than 6 years after their surgery. Concomitant CABG surgery does not diminish median survivorship among patients >80 years of age.

Outcomes of patients undergoing concomitant aortic and mitral valve surgery in northern new England.

BACKGROUND: Concomitant aortic (AV) and mitral (MV) valve surgery accounts for 4% of all valve procedures in northern New England. We examined in-hospital and long-term mortality. METHODS AND RESULTS: This is a report of a prospective study of 1057 patients undergoing concomitant AV and MV surgery from 1989 to 2007. The Social Security Administration Death Master File was used to assess long-term survival. Kaplan-Meier and log-rank tests were performed. In-hospital mortality was 15.5% (11.0% for patients <70 years, 18.0% for 70- to 79-year-olds, and 24% for those > or =80 years). Overall median survival was 7.3 years. Median survival without coronary artery bypass grafting was 9.5 years and with coronary artery bypass grafting was 5.7 years (P<0.001). Survival in women was worse than in men (7.3 versus 9.3, years, P=0.033). Median survival by age was 11.0 years for patients <70 years, 5.4 years for 70- to 79-year-olds, and 4.8 years for those > or =80 years. Median survival was not significantly different for patients > or =80 years compared with those who were 70 to 79 years old (P=0.245). CONCLUSIONS: Double-valve surgery has a high in-hospital mortality rate and a median survival of 7.3 years. After patients have survived surgery, long-term survival is similar between men and women, smaller and larger patients, and those receiving MV repair or replacement. Survival continues to decline after surviving surgery for patients > or =70 years old and those who undergo concomitant coronary artery bypass grafting. In patients <70 years, either mechanical valves in both positions or a tissue AV and mitral repair have the lowest in-hospital mortality and the best long-term survival. In patients > or =70 years, tissue valves in both positions have the best in-hospital and long-term survival.

The association of perioperative red blood cell transfusions and decreased long-term survival after cardiac surgery.

BACKGROUND: Exposure to red blood cell (RBC) transfusions has been associated with increased mortality after cardiac surgery. We examined long-term survival for cardiac surgical patients who received one or two RBC units during index hospitalization. METHODS: Nine thousand seventy-nine consecutive patients undergoing coronary artery bypass graft, valve, or coronary artery bypass graft/valve surgery at eight centers in northern New England during 2001-2004 were examined after exclusions. A probabilistic match between the regional registry and the Social Security Administration's Death Master File determined mortality through June 30, 2006. Cox Proportional Hazard and propensity methods were used to calculate adjusted hazard ratios. RESULTS: Thirty-six percent of patients (n = 3254) were exposed to one or two RBC units. Forty-three percent of RBCs were given intraoperatively, 56% in the postoperative period and 1% were preoperative. Patients transfused were more likely to be anemic, older, smaller, female and with more comorbid illness. Survival was significantly decreased for all patients exposed to 1 or 2 U of RBCs during hospitalization for cardiac surgery compared with those who received none (P < 0.001). After adjustment for patient and disease characteristics, patients exposed to 1 or 2 U of RBCs had a 16% higher long-term mortality risk (adjusted hazard ratios = 1.16, 95% CI: 1.01-1.34, P = 0.035). CONCLUSIONS: Exposure to 1 or 2 U of RBCs was associated with a 16% increased hazard of decreased survival after cardiac surgery.

Prediction of survival after coronary revascularization: modeling short-term, mid-term, and long-term survival.

BACKGROUND: Many clinical prediction rules for short-term mortality after coronary revascularization have been developed, validated, and introduced into routine clinical practice. Few rules exist for predicting long-term survival in the modern era of coronary revascularization. We report on the development and validation of models for predicting long-term survival after coronary artery bypass graft surgery and percutaneous coronary intervention on the basis of recent experience. METHODS: We linked 1987 through 2001 coronary artery bypass graft surgery and 1992 through 2001 percutaneous coronary intervention data from our northern New England registries on 35,000 patients with complete data on risk factors to the National Death Index, ascertaining 7,000 deaths. We partitioned time after revascularization into three periods on the basis of exploratory analysis using generalizations of Cox's semiparametric model to nonproportional hazards and nonlinear log-hazards. These periods were 0 to 3 months, 4 to 18 months, and 19 months and later. For each period, Cox's regression model was used to regress survival on risk factors yielding three models, which were then combined to make long-term predictions. RESULTS: These models were incorporated into easy-to-use software that yields predicted survival for up to 8 years after revascularization. The Harrell concordance statistic ranged from 72% to 83% for these models. CONCLUSIONS: We developed and internally validated models that accurately predict long-term survival after coronary artery bypass graft surgery and percutaneous coronary intervention as currently performed. These models using routine clinical data, can be solved with available software, and could be used to enhance informed, patient-centered clinical decision making on the choice of revascularization procedure.

Appropriateness of coronary artery bypass graft surgery performed in northern New England.

OBJECTIVES: The goal of this study was to assess the concordance between the American College of Cardiology (ACC) and the American Heart Association (AHA) 2004 Guideline Update for Coronary Artery Bypass Graft Surgery and actual clinical practice. BACKGROUND: There is substantial geographic variability in the population-based rates of coronary artery bypass graft (CABG) procedures, and in recent years, there have been several public concerns about unnecessary cardiac care. The actual rate of inappropriate cardiac procedures is unknown. METHODS: We evaluated 4,684 consecutive isolated coronary artery bypass graft procedures performed in 2004 and 2005 in northern New England. Our regional registry data were used to categorize patients into clinical subgroups. Detailed clinical criteria were then used to categorize procedures within these subgroups as class I (useful and effective), class IIa (evidence favors usefulness), class IIb (evidence less well established), and class III (not useful or effective). RESULTS: Among these 4,684 procedures, we were able to classify 99.6% (n = 4,665). The majority of procedures were class I (87.7%). Class II procedures totaled 10.9%. The remaining 1.4% of procedures were class III. CONCLUSIONS: In this regional study, we found that 98.6% of CABG procedures that could be classified were considered to be appropriate. In these data, actual clinical practice closely follows the recommendations of the 2004 ACC/AHA guidelines for CABG surgery.

Cardiopulmonary bypass recommendations in adults: the northern New England experience.

Using a regional cardiopulmonary bypass (CPB) registry, we compared the practice of CPB at eight northern New England institutions to recently published recommendations. We examined CPB practice among 3597 adult patients undergoing isolated coronary artery bypass grafting surgery from January 2004 to June 2005. Registry variables were used to compare regional CPB practice to recommendations on topics of neurologic protection (pH management, avoidance of hyperthermia, minimizing return of pericardial suction blood, aortic assessment, arterial line filtration), maintenance of euglycemia, reduction of hemodilution, and attenuation of the inflammatory response. We report overall regional practice (regional minimum, maximum). All centers used alpha-stat pH management and arterial line filters. Avoidance of hyperthermia (temperature < 37degrees C) was achieved during 23.4% of procedures (regional minimum, 1.5%; maximum, 83.2%). Minimizing return of pericardial suction blood was achieved in 23.7% of cases (0.7%, 93.6%). Aortic assessment was performed during 45.7% of procedures (1.3%, 98.9%). Maintenance of euglycemia (< 200 mg/dL) was accomplished in 82.7% (57.1%, 97.9%) of cases. Hemodilution (hematocrit < 23% on CPB) was lower for men 32.4% (20.6%, 52.3%) than women 77.9% (64.7% 88.9%). Men were less likely to receive red blood cell transfusions in the operating room (11.0%; 1.8%, 20.9%) than women (54.6%; 30.1%, 70.6%). In an effort to attenuate the inflammatory response, surface coated circuits were used in 83.3% of procedures (8.8%, 100%). During this time, gaps existed between regional CPB practice and recently published recommendations. We continue to prospectively measure CPB practice relating to these recommendations to monitor and improve the care provided to our patients.

Long-term survival of the very elderly undergoing coronary artery bypass grafting.

BACKGROUND: Increasing numbers of the very elderly are undergoing coronary artery bypass graft surgery (CABG). Short-term results have been studied, but few data are available concerning long-term outcomes. METHODS: We conducted a cohort study of 54,397 consecutive patients undergoing primary, isolated CABG surgery between July 1, 1987, and June 30, 2006. Patient records were linked to the Social Security Administration's Death Master File. RESULTS: During 390,871 person-years of follow-up, there were 17,352 deaths. There were 51,149 patients younger than 80 years, 2,661 patients aged 80 to 84 years, and 587 patients aged 85 or more years who underwent isolated CABG surgery. Crude in-hospital survival was 97.2% for those less than 80 years, 98.3% for those aged 80 to 84 years, and 87.6% for those aged 85 or more years. Patients aged 80 or more years were more likely to be female (46.9%), more likely to be emergency priority (10.2%), and more likely to have associated comorbidities than younger patients. Patients aged 85 or more years were more likely to have intraoperative and postoperative morbid events. Among patients younger than 80, median survival was 14.4 years with an annual incidence of death of 4.2%. Among patients 80 to 84 years old, median survival time was 7.4 years, with an annual incidence rate of death of 10.3%. Among patients aged 85 or more years, median survival was 5.8 years, and the annual incidence of death was 13.7%. CONCLUSIONS: Although very elderly CABG patients have more comorbidities and more acute presentation than younger patients and their in-hospital mortality rate is high, their long-term survival is surprisingly good.