Over the past few decades, the risk profile of patients referred for isolated coronary artery bypass graft (CABG) surgery has changed considerably.1-3 Historical data supporting the notion that patients undergoing surgical revascularisation today are much older and sicker have come from single and multi-institution studies.1-5 CABG surgery is increasingly being performed on older people, including octogenarians with multiple comorbidities.6,7

In addition, the number of patients undergoing percutaneous coronary intervention (PCI) has risen dramatically with advancements in the technology of interventional cardiology, including angioplasty and stent deployment.8,9 PCI has not only been used in low-risk patients with straightforward lesions but is increasingly being attempted in high-risk patients with multivessel coronary lesions.10

However, surgical revascularisation has benefits that include a significant reduction in 30-day operative mortality rates, despite a substantial increase in preoperative comorbidities of isolated CABG patients,2,3 indicating a major improvement in outcomes from cardiac surgery. About 13 000 CABG operations are performed annually in Australia.

Until the Australasian Society of Cardiac and Thoracic Surgeons (ASCTS) database registry was established in 2001, there was no opportunity to monitor trends in preoperative risk and postoperative outcomes for patients undergoing CABG surgery in Australia. Here, we investigate changes in the preoperative clinical characteristics of this population and their major postoperative morbidities for the 5 financial years from 2001–02 to 2005–06.

A total of 14 526 cardiac surgical procedures were conducted in Victorian public hospitals between July 2001 and June 2006. CABGs were consistently the most commonly conducted procedure, with 9372 procedures (65%) being isolated CABGs (92% performed with cardiopulmonary bypass).

Demographic and comorbidity details for patients undergoing isolated CABG surgery are shown in Box 1 and Box 2 for each financial year, to compare yearly trends in patient characteristics.

There was an increase in the mean age of patients undergoing CABG surgery over the observation period, from 65.4 years in 2001–02 to 66.0 years in 2005–06 (P < 0.001). Patients throughout the study period were predominantly male (range, 75.2%–79.0%). Increases were seen in the proportion of patients with hypertension (P < 0.001), respiratory disease (P < 0.001), and previous cardiac history including left main coronary artery stenosis > 50% (P = 0.03) and recent myocardial infarction occurring 1–21 days before surgery (P < 0.001). The number of patients undergoing repeat CABG surgery decreased from 4.4% to 2.6% (P = 0.002). Non-elective surgery accounted for 38% of total CABG surgery in 2001–02 but increased to 46% in 2005–06 (P < 0.001).

The data showed no significant changes in other risk factors, including diabetes, hypercholesterolaemia, cerebrovascular disease, peripheral vascular disease, renal dysfunction, congestive heart failure, unstable angina, arrhythmia, cardiogenic shock, triple vessel disease and severe left ventricular impairment.

The 30-day mortality rate for each financial year is shown in Box 3. The patients’ AusSCORE-calculated preoperative predicted risk for isolated CABG surgery remained unchanged over the 5 years (1.7% to 1.8%). The observed mortality rate within 30 days of CABG surgery declined from 2.2% in 2001–02 to 1.8% in 2005–06, but this change was not statistically significant (P = 0.983). There was a non-significant 50% decrease in mortality for elective procedures, from 1.2% in 2001–02 to 0.6% in 2005–06 (P = 0.151), while the 30-day mortality rate for non-elective surgery ranged between 2.9% and 3.7% over the 5 years; this trend was also not statistically significant (P = 0.977).

Postoperative morbidity trends for CABG surgery remained steady over the 5 years (Box 3).

The main finding from this study is that the 30-day mortality rate remained unchanged over the 5 years, despite an increase in older patients undergoing isolated CABG surgery. The unadjusted mortality rate of 1.8% for the study period is the same as the 30-day mortality rate reported in the United Kingdom and is lower than the United States figure of 2.3%.1,14 Postoperative complications after CABG surgery also remained steady and consistently low over the study period and compared well with international rates.4,15 The data suggest that 30-day outcomes after CABG surgery in Victoria are favourable compared with any population undergoing this surgery.

We found that the mean age of patients undergoing isolated CABG surgery in Victorian public hospitals increased over the study period, reflecting other studies’ findings that older patients are being increasingly referred for surgical coronary artery revascularisation.6,7 Benefits of CABG surgery in older people include demonstrated improvement in functional status.16

Interestingly, the frequency of repeat CABG surgery declined, possibly due to the introduction of coronary stenting or to improved medical management.8,9 Although studies have reported no differences in mortality between multivessel stenting and CABGs, high rates of reintervention with repeat revascularisation being needed after PCI continue to be a problem.10

We found an increased incidence of some cardiovascular risk factors and increasing acuity of patients presenting for surgical treatment, along with an associated increase in non-elective surgery. Despite these increases, the trend in 30-day mortality rates in patients undergoing non-elective surgery was not statistically significant.

The AusSCORE predictive mortality outcome model was developed and has been deemed suitable for use in the Australian context.13 It has demonstrated accuracy in predicting risk — the observed/expected outcome ratio ranges between 0.9% and 1.2%. Minimal change in predicted mortality was observed in this cohort, presumably because only a few of the risk factors used in the model increased over the study period (see Box 1 and Box 2).

The ASCTS database was established for quality control and improvement purposes, and its strength lies in the inclusion of data from all six Victorian public hospitals that perform CABG surgery. The quality of data collected has been high and, importantly, with very few missing data (less than 1%). This ensures an accurate presentation of trends in CABG surgery. It also allows valid performance indicators to be established and assists with forecasting trends in the population presenting for CABG surgery.

Implementation of routine data collection and reporting programs for quality improvement, including the ASCTS and other large international cardiac surgery databases, has improved reported operative mortality rates after CABG surgery.1,3,15 Evidence from these databases suggests that feedback on benchmarks and performance indicators can favourably impact on the outcomes after CABG surgery. In addition, technical advances in cardiac surgery and cardiology, with the introduction of new pharmacological agents, techniques and devices, have led to reduced mortality over the past decade. Interestingly, the authors of one study that reported reduced rates of hospitalisation for acute myocardial infarction and revascularisation (PCI and cardiac surgery) believed better control of risk factors was one explanation for the reduction,17 although they questioned the sensitivity of their database.

Some limitations of our study need to be acknowledged. First, at the time of the study, the ASCTS database was in use in only one Australian state and therefore did not represent a comprehensive national experience. However, the ASCTS database has since expanded into a national program, with the recent participation of 13 cardiac units from four other states. This now represents over 70% of public cardiac surgical units in Australia.

Second, the database only collects information on inhospital postoperative morbidities and does not capture events that occur after discharge from hospital, which may represent a significant proportion of all events. It is therefore likely to underestimate the real burden of complications, given the number of comorbid conditions in the cohort of patients presenting for surgery.

Further expansion of the database to incorporate other states will allow for accurate comparisons and identification of risk factors associated with complications, and the development of models for all types of cardiac surgery. We are in the process of developing a web-based system that will streamline data collection and provide real-time benchmarking measures. Although understanding short-term mortality and complications is important, the next step is to follow up these patients for several years to examine the outcomes of CABG surgery over time.