Operative mortality after coronary artery bypass graft surgery

INTRODUCTION — The treatment of coronary heart disease has evolved significantly due in part to improvements in both medical therapy and surgical and percutaneous revascularization techniques. The majority of patients with chronic stable angina are treated with medical therapy, but there are a variety of indications for coronary artery bypass graft surgery (CABG) or percutaneous coronary intervention. (See "Chronic coronary syndrome: Indications for revascularization".)

The perioperative and in-hospital mortality rate after CABG will be reviewed here. Cardiac and noncardiac complications of CABG and the long-term outcome after CABG are discussed separately. (See "Early cardiac complications of coronary artery bypass graft surgery" and "Early noncardiac complications of coronary artery bypass graft surgery" and "Coronary artery bypass graft surgery: Graft choices".)

GENERAL PREVENTIVE THERAPY — The 2011 American College of Cardiology/American Heart Association guideline for CABG issued general recommendations for preventive measures to minimize the risk of both morbidity and mortality after CABG [1]:

●Aspirin to improve morbidity and mortality

●Beta blockers to prevent perioperative atrial fibrillation; amiodarone and sotalol are alternatives

●Statin therapy

●Prophylactic antimicrobials to prevent surgical site infection

●Glycemic control (using an insulin infusion) during the perioperative period

The evidence supporting the efficacy of these therapies is presented separately. (See "Coronary artery bypass surgery: Perioperative medical management" and "Atrial fibrillation and flutter after cardiac surgery", section on 'Prevention of atrial fibrillation and complications'.)

In addition to these medical therapies, carotid duplex ultrasound may be performed to identify severe carotid stenosis in patients with an audible bruit, severe peripheral artery disease, or a previous stroke or transient ischemic attack. (See "Neurologic complications of cardiac surgery", section on 'Risk factors'.)

OPERATIVE MORTALITY — Using registry data in the United States, the perioperative and in-hospital mortality rate after coronary artery bypass graft surgery (CABG) between 1997 and 2001 averaged about 1 percent for the lowest risk elective patients, and 2 to 5 percent for all patients [2-4]. However, the risk is highly dependent upon comorbid disease, postoperative complications, and the hospital volume of CABG procedures [3-6]. Patients who undergo CABG at low volume centers are more likely to be at high risk [5], while patients who undergo CABG at high volume centers are more likely to be at low risk [6].

These mortality rates are probably not applicable to all countries. This was illustrated in a report from the National Health Service in Brazil in which the perioperative mortality was 7.0 percent in almost 66,000 CABG procedures performed between 2000 and 2003 [7]. Possible contributors to this increase in risk are bias (some of the American registries are voluntary and therefore may underestimate events) and a greater number of cases performed in low-volume hospitals. The difference in outcomes may affect clinical decision making on the indications for CABG.

PREDICTORS OF RISK — A number of preoperative and intraoperative factors have been demonstrated to influence the rate of operative mortality after coronary artery bypass graft surgery (CABG).

Risk prediction algorithms — The most effective way of stratifying cardiac surgical patients according to operative risk is to use one of several available risk prediction algorithms that incorporate multiple variables to derive a risk score. One of the most widely used algorithms for this purpose in the United States is that developed by the Society of Thoracic Surgeons (STS) [8,9]. This algorithm, which incorporates patient age, sex, and variables reflecting comorbidities as well as the severity and acuity of cardiac disease, is proprietary and available only to institutions participating in the STS database [10]. Another such algorithm, the EuroSCORE (table 1), is widely used in Europe [11,12], is available and has been validated in a North American population [13].

One analysis compared the performance of the STS algorithm and the EuroSCORE in predicting the survival of 4497 patients undergoing isolated CABG between 1996 and 2001 [14]. Overall 30-day mortality was 1.9 percent and was accurately predicted by both risk scores. Within individual deciles of risk, both scores performed well, but the EuroSCORE had significantly greater discriminatory power.

A variety of other algorithms are also available for patients undergoing CABG. In a review of 19 such algorithms in 4351 patients undergoing isolated CABG, EuroSCORE had the highest discriminatory power for both 30-day and one-year mortality, followed by algorithms from New York State and the Cleveland Clinic [15].

Limitations of the models — The principle limitation of these risk prediction algorithms is that they may not be applicable to patient populations, which differ in one or more significant ways [16]. As the patient population changes and as surgical technique changes, risk analyses can become outdated. Thus, the greater discriminatory power of the EuroSCORE in the above analysis may derive, at least in part, from the fact that the patient data upon which it is based were collected in 1995, while those upon which the STS algorithm is based were collected from 1990 to 1993. The STS continues to gather new data to permit progressive algorithm revision.

Another example of this limitation is that currently, most patients referred for CABG undergo an elective procedure, but that the risk models discussed above were derived from patients who underwent both elective and urgent/emergent procedures. The ACEF (Age, Creatinine, Ejection fraction) score was derived to address this issue, as it included only three variables and was developed from a population of individuals who underwent elective open heart surgery at a single institution [17]. ACEF was able to predict mortality as well as or better than five other more complex risk scores such as EUROSCORE or the Cleveland Clinic model.

A related limitation is that, as many patients are candidates for either CABG or percutaneous coronary intervention (PCI), a simple, bedside risk prediction model that could be applied to either procedure would be useful. The Mayo Clinic Risk Score (MCRS), which uses seven preprocedural variables, is the first such model. It was initially validated in a population of patients scheduled to undergo PCI [18]. (See "Periprocedural complications of percutaneous coronary intervention".)

When tested in over 300,000 patients scheduled to undergo isolated CABG surgery in the STS database, there was moderate positive correlation between MCRS and the observed in-hospital mortality [19]. When compared to the 26 variable STS algorithm, performance was inferior.

Hospital and surgeon experience — The in-hospital mortality after CABG is related to the volume of CABG procedures performed at the hospital, as illustrated by the following observations:

●In an analysis of over 900,000 CABG procedures in the Medicare database between 1994 and 1999, the adjusted in-hospital or 30-day mortality was inversely related to hospital volume, ranging from 6.1 percent for hospitals performing <230 procedures per year to 4.8 percent when the volume was >849 procedures per year [3].

●In a report on over 200,000 procedures from the STS National Cardiac Database between January 2000 and December 2001, the overall operative mortality was 2.7 percent [4]. The mortality rate ranged from 3.5 percent for hospitals performing ≤150 procedures per year to 2.4 percent for hospitals performing >450 procedures per year.

●Similar observations have been made in Brazil where a report from the National Health Service for all types of cardiac surgery noted a progressive increase in perioperative mortality from 5.8 percent in hospitals performing more than 341 procedures per year (highest quartile) to 9.7 percent in hospitals performing less than 131 procedures per year (lowest quartile) [7].

The association between volume and operative mortality may depend upon patient risk status:

●In the STS database report, adjustment for clinical risk, year of surgery, and patient clustering within centers reduced the impact of annual hospital volume on mortality, with an adjusted mortality range from 3.1 percent (≤150 procedures) to 2.4 percent (>450 procedures) [4]. For patients under 65 years and for those estimated to be at low operative risk, there was no association of hospital volume with mortality.

●Similar findings were noted in a review of 13,644 patients seen in hospitals performing a low (<200) or high (>200) number of procedures per year: Four established predictors (age, sex, surgical priority, and severity of illness) were used to classify risk status (minimal, low, moderate, high, and severe) [20]. In-hospital mortality in low-volume hospitals was significantly higher in patients at moderate risk (5.3 versus 2.2 percent for high-volume hospitals) and high risk (23 versus 12 percent); there was no difference in mortality in the other risk groups.

●A different conclusion, however, was reached in an analysis of 57,150 CABG operations performed in New York State between 1997 and 1999; there were 1260 (2.2 percent) in-hospital deaths [21]. A logistic regression model was used to classify patient risk as low (predicted in-hospital mortality <2 percent) or moderate to high (predicted in-hospital mortality ≥2 percent). When low-volume hospitals were defined as those performing fewer than 200 operations per year, the odds of in-hospital death were lower at the high-volume hospitals for both low- and moderate- to high-risk patients (adjusted odds ratios 0.53 and 0.62, respectively). The advantage of high-volume hospitals persisted when the definition of low volume was changed to less than 600 operations per year.

The experience of the individual surgeon is a predictor of mortality independent of the hospital volume:

●In a study of over 200,000 CABG procedures in the Medicare database from 1998 and 1999, operative mortality rates for surgeons performing <101, 101 to 162, and >162 CABGs per year were 5.4, 4.3, and 4.0 percent, respectively (odds ratio 1.36). This relationship remained significant after adjusting for hospital volume (odds ratio 1.33).

●In the STS database analysis, surgeon volume was also a determinant of risk-adjusted mortality [4]. Surgeons performing 10 to 85, 86 to 138, and ≥139 CABG procedures annually had overall mortality rates of 2.9, 2.6, and 2.4 percent, respectively. The highest mortality rates (3.3 percent) were observed when patients were treated by low-volume surgeons at low-volume hospitals.

●In the New York State study, surgeon volume was also a determinant of mortality for both low- and moderate- to high-risk patients [21]. For low-risk patients, either high surgeon volume or high hospital volume or both significantly reduced in-hospital mortality. For moderate- to high-risk patients, only the combination of both high surgeon volume and high hospital volume was associated with a significant reduction in risk.

Hospital volume and the experience of the surgeon also affect the likelihood of early readmission after CABG. (See "Early noncardiac complications of coronary artery bypass graft surgery", section on 'Early readmission'.)

Left ventricular dysfunction — The presence of left ventricular dysfunction and heart failure is one of the most important independent predictors of operative mortality and other major adverse events after CABG [22,23]. In a prospective observational study of over 8600 patients undergoing CABG between 1992 and 1997, the operative mortality varied from less than 2 percent with a left ventricular ejection fraction (LVEF) >40 percent to 3.5 to 4 percent with an LVEF between 20 and 40 percent to approximately 8 percent with an LVEF <20 percent [23].

Increasing age — Not surprisingly, older age is a predictor of increased risk after CABG [22,24]. This is an important issue since an increasing number of elderly patients (ie, ≥70 years of age) are undergoing CABG.

One report retrospectively reviewed the outcome of 3330 patients over the age of 70 in whom isolated CABG was performed between 1982 and 1996 [25]. The overall operative mortality decreased from 7.2 to 4.4 percent over the course of the study. The prevalence of high-risk older patients rose from 16.2 to 26.9 percent, but the operative mortality in these individuals fell from 17.2 to 8.9 percent. Independent predictors of mortality in older adults were similar to those of younger patients and included poor left ventricular function, previous CABG, female sex, peripheral artery disease, and diabetes.

A later report evaluated the outcome of 725 patients ≥80 years of age who underwent CABG between 1996 and 2001 in Canadian centers [26]. By 2001, 6.2 percent of bypass operations were performed in octogenarians. Compared to younger patients, octogenarians have significantly higher rates of in-hospital mortality (9.2 versus 3.8 percent) and stroke (4.7 versus 1.6 percent) after CABG. After adjustment of comorbidities, age ≥80 was an independent predictor of mortality. The relative increase in mortality was greatest with urgent (10.3 versus 3.9) and emergency surgery (17.0 versus 12.1 percent).

Acute kidney injury — In a study of 3460 adults, the in-hospital mortality rate increased progressively with greater reductions in estimated glomerular filtration rate (GFR; 1.0, 10, 29, and 39 percent for patients with reductions in estimated glomerular filtration rate of ≤25 percent, >25 percent or dialysis, >50 percent or dialysis, or >75 percent or dialysis, respectively) [27].

Chronic kidney disease — The EuroSCORE risk prediction algorithm includes underlying chronic kidney disease (serum creatinine >2.3 mg/dL [200 µmol/L]) as a risk factor for surgical mortality (table 1) [11-13]. A subsequent prospective study of 4403 consecutive patients undergoing first-time isolated CABG found that even mild renal dysfunction (serum creatinine 1.5 to 2.3 mg/dL [133 to 199 µmol/L]) was associated with a higher operative mortality than in patients with serum creatinine concentrations under 1.5 mg/dL (6.1 versus 2.1 percent, multivariate adjusted odds ratio 1.91, 95% confidence interval [CI] 1.18-3.03) [28].

A graded relationship between the degree of renal dysfunction and operative mortality was noted in a report from the Society of Thoracic Surgeons National Adult Cardiac Database of almost 500,000 patients undergoing isolated CABG [29]. Operative mortality rose from 1.3 percent in patients with a normal estimated GFR (≥90 mL/min) to 1.8 percent with a GFR of 60 to 89 mL/min (mild kidney disease, present in 51 percent), 4.3 percent with a GFR of 30 to 59 mL/min (moderate kidney disease, present in 24 percent), and 9.0 to 9.3 percent with a GFR less than 30 mL/min or on dialysis (severe kidney disease, present in 4 percent). After multivariable adjustment, preoperative GFR was one of the most powerful predictors of operative mortality (adjusted odds ratio for mortality 1.55 and 3.82 with moderate and severe kidney disease, respectively).

In addition to the increase in operative mortality and morbidity, chronic kidney disease is also predictor of increased long-term mortality after CABG [30]. This finding is consistent with the long-term increase in cardiovascular risk associated with chronic kidney disease. (See "Chronic kidney disease and coronary heart disease".)

New Q waves — New Q waves developing after CABG were associated with increased in-hospital mortality in a report from the Coronary Artery Surgery Study (9.7 versus 1.0 percent in those without new Q waves) [31]. Although usually reflecting a new myocardial infarction (MI), new Q waves after CABG can also represent unmasking of a previous MI [32].

Single or multiple IMA grafts — The use of the left internal mammary artery (LIMA) is associated with a higher graft patency rate and a lower perioperative mortality, similar to the known benefit of an arterial graft on long-term outcome compared to saphenous vein grafts (SVGs) (figure 1 and figure 2) [33,34]. One study of 21,873 patients who underwent an isolated, first-time CABG found that the adjusted odds ratio for in-hospital mortality with the use of a LIMA, compared to no LIMA, was 0.40 [35]. LIMA grafts were protective across all major patient and disease groups.

Because of the better outcomes with arterial compared to SVGs, there is increasing use of multiple arterial grafts in patients with multivessel disease. In-hospital mortality appears to be similar with multiple and single arterial grafts [36-38].

Coronary artery diameter — Perioperative mortality is increased with smaller coronary artery diameters, which may explain the increased risk that has been observed in women, particularly those <50 years of age, and smaller people [39-41]. It is possible that smaller coronary arteries are more likely to be associated with thrombosis, technical difficulties, and reduced short-term graft patency. These issues were evaluated in a prospective study of 1325 patients undergoing CABG: Small mid left anterior descending artery diameter was associated with an increase in in-hospital mortality [40]:

●1.0 mm diameter – 15.8 percent

●1.5 to 2.0 mm diameter – 4.6 percent

●2.5 to 3.5 mm diameter – 1.5 percent

Vessel size was strongly correlated with sex and body size measures (body surface area, body mass index, height, and weight). After adjusting for age and body size, sex remained an important predictor of coronary artery size.

Gastrointestinal complications — Gastrointestinal (GI) complications after CABG are infrequent but are associated with increased mortality. The largest reported experience comes from the United States Agency for Healthcare Research and Quality, which assessed the incidence and impact of GI complications after 2.7 million CABG operations performed from 1998 to 2002 [42]. The following observations were noted:

●The incidence of GI complications was 4.1 percent

●The inpatient mortality was significantly increased in those with GI complications compared to those without (12.0 versus 2.5 percent).

(See "Early noncardiac complications of coronary artery bypass graft surgery", section on 'Gastrointestinal'.)

Metabolic syndrome — The metabolic syndrome, which is a risk factor for cardiovascular disease, is common in patients who undergo CABG. In an observational study of 5304 patients who underwent an isolated CABG, the metabolic syndrome (present in 46 percent of patients) was a significant independent predictor of operative mortality (relative risk 3.04, 95% CI 1.73-5.32) [43]. (See "Metabolic syndrome (insulin resistance syndrome or syndrome X)", section on 'Clinical evaluation and implications of diagnosis'.)

Preoperative anemia — Preoperative anemia is an independent risk factor for early mortality after CABG; three large observational studies support this point [44-46]. For example, the Multicenter Study of Preoperative Ischemia Epidemiology II evaluated the relationship between the lowest preoperative hemoglobin concentration and adverse postoperative outcomes in over 4800 patients undergoing elective CABG surgery [44,47]. Preoperative anemia (<13 g/dL in men and <12 g/dL in women) was present in 28 and 36 percent of patients, respectively. CABG patients with preoperative anemia (starting at hemoglobin concentrations <11 g/dL) were at increased risk of postoperative cardiac and noncardiac events in a dose-dependent manner.

Duration of red cell storage — Longer red blood cell storage times are associated with structural and functional changes that may worsen outcomes in the recipient. In a retrospective study of over 6000 patients who underwent open heart surgery and were transfused, those who received blood stored for more than 14 days had significantly worse outcomes (intubation beyond 72 hours, renal failure, sepsis or septicemia, and in-hospital and one-year mortality) compared to those with shorter storage times [48].

However, other studies have not demonstrated an independent effect of red cell storage time on outcomes after CABG and any potential benefits of using blood stored for two weeks or less must be weighed against the difficulty of maintaining the blood supply if only fresher units are used for transfusion. These issues are discussed in detail separately. (See "Practical aspects of red blood cell transfusion in adults: Storage, processing, modifications, and infusion", section on 'RBC age/storage duration effect on clinical outcomes'.)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, “The Basics” and “Beyond the Basics.” The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on “patient info” and the keyword(s) of interest.)

●Beyond the Basics topics (see "Patient education: Coronary artery bypass graft surgery (Beyond the Basics)" and "Patient education: Recovery after coronary artery bypass graft surgery (CABG) (Beyond the Basics)")

SUMMARY

●The perioperative mortality for patients undergoing coronary artery bypass graft surgery (CABG) varies according to the extent of patient comorbidities. Mortality is about 1 percent for the lowest-risk elective patients and 2 to 5 percent for all patients. (See 'Operative mortality' above.)

●A number of preoperative and intraoperative factors have been demonstrated to influence the rate of operative mortality after CABG. These include hospital and surgeon experience, degree of left ventricular systolic dysfunction, increasing age, the presence and extent chronic kidney disease, and the type and number of bypass grafts. (See 'Predictors of risk' above.)

●Risk prediction algorithms have been developed to predict perioperative risk. EuroSCORE and the Society of Thoracic Surgeons algorithm are two widely used examples. (See 'Risk prediction algorithms' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff thank Dr. Julian M. Aroesty for his contributions as an author to prior versions of this topic review.