Heart Care Info - Heart Disease Prevention & Treatment

The 4th European Association for Cardio-Thoracic Surgery adult cardiac surgery database report

B Bridgewater, R Kinsman, P Walton... - Interactive ..., 2011 - Eur Assoc Cardio Surg The EACTS annual meeting in Geneva this September saw the release of the 4th EACTS adult cardiac surgery database report [1]. It contains an analysis of over 1,000,000 patients from 366 hospitals in 29 countries. Most submitting countries are in Europe, but we also have ...

The EACTS annual meeting in Geneva this September saw the release of the 4th EACTS adult cardiac surgery database report [1]. It contains an analysis of over 1,000,000 patients from 366 hospitals in 29 countries. Most submitting countries are in Europe, but we also have important contributions from China including Hong Kong. It is a voluntary database and some countries submit data on all cardiac surgery operations undertaken, others only for surgery performed in a smaller numbers of hospitals. It is testament to the professional societies representing cardiac surgery in these countries, along with the hospitals and surgeons involved, that data from so many patients have been collected in registries and further tribute to the trust placed in EACTS and their partners that these data have been returned and permissions granted for analysis.

The report gives many interesting findings, including variation in the proportion of cardiac surgery that is isolated coronary artery bypass surgery (CABG), with an overall proportion of 56.8% in the database, ranging from 29.7% to 79.6% across countries. Over time there are trends towards a decreasing proportion of CABG along with an increase in valve surgery for the larger contributing countries. The majority of countries have returned mortality data for analysis, and this is close to complete in 18 countries and complete in more than 95% of patients for a further three. Four countries have submitted no mortality data. Overall mortality for isolated CABG is 2.2% (n=219,053), for isolated valve surgery 3.7% (n=75,247) and for combined valve and CABG 6.2% (n=37,721). These data from so many patients should act as useful crude contemporary benchmarks.

An analysis of the logistic Euro SCORE confirms poor predictive ability for contemporary practice, suggesting that the model should not be used for giving predicted mortality in an unadjusted form [2]. This leaves an urgent need for better models to support quality improvement, quality assurance purposes, patient consent and support for multi-disciplinary decision making.

There are extensive comparisons included about the incidence of risk factors for patients coming to CABG between countries and many of these, such as the incidence of women undergoing surgery and the proportion of operations performed as non-elective surgery raise important questions about the comparative configuration of, and access to, cardiac surgery. We have included extensive summary data on populations, wealth and burden of cardiovascular disease from the World Health Organisation to help see these data in context. We hope that the information presented will stimulate further local consideration of these issues. Along a similar theme we have included analysis of postoperative length of in-hospital stay (LOS) following isolated CABG and explored associations between risk factors and LOS. Overall median LOS is seven days, but varies from 4.5 days up to 11 days between countries. We have listed these data and would again hope that this information would be used locally to understand variance and drive quality.

The EACTS report gives an opportunity to analyse current practice of on-pump vs. off-pump surgery, and the use of venous or arterial grafts. The overall proportion of isolated CABG performed off pump is 20.2% and varies from 0.8% up to 91.4%, between countries. Mortality for on-pump surgery is higher, but we know there are differences between the groups with respect to case mix. Overall, 20.0% of patients receive more than one arterial graft. This varies from 1.2% to 76.6% between countries.

Previous EACTS reports have looked at CABG in some detail, but the quality of submitted data has now improved to allow an analysis of aspects of valve surgery, including 62,545 isolated aortic valve (AV) operations and 32,880 combined AV and CABG operations. This shows the incidence of various risk factors and their association with mortality and of particular note in the era of novel approaches to AV surgery are data that show that more than 20% of conventional AV surgery is performed on patients over 80 in some countries. Overall mortality for isolated AV surgery for patients under 56 is 1.2% and for those over 80 it is 6.1%. We have analysed the choice of prosthesis for AV replacement for those countries that have returned the required data (comprising 148,376 operations) and demonstrated marked increase in the use of biological valves over time, for all age groups. The proportion of patients aged between 61 and 70 years of age undergoing isolated AV replacement with biological valves has increased from 49.2% to 73.1% between 2003 and 2008.

We have included an analysis in the report of 15,075 patients undergoing mitral valve (MV) surgery, and in many ways this section raises more questions than it provides answers. The incidence of missing data on valve pathology is high, and drawing conclusions about mitral surgery in the absence of knowledge of whether the valve is rheumatic, degenerative or ischaemic is impossible. However, this is an important finding in its own right: for us to maximize the value of MV registry data, they must be more complete. We would hope that hospitals and professional societies would respond to this issue, and have given the opportunity for countries to understand their data quality completeness rates throughout the report. However, despite these caveats, we have still seen marked variations between countries in the proportion of patients coming to MV surgery with significant impairment of left ventricular function, and NYHA class 3–4 symptoms, which we hope will be significant drivers to deeper understanding and improvement. There is also variation in the ratio of MV repair to valve replacement, which again needs to be interpreted against local understanding of valve pathology. As with previous studies, the mortality for mitral repair is significantly lower in the registry, particularly for the elderly, at 1.4% for isolated repair against 1.6% for replacement for those under 56, rising to 5.7% and 11.9%, respectively for those over 80 years of age.

The latest database report is significant in view of the large volume of patients submitted from so many countries. The database is an amalgamation of multiple registries, and unlike a randomised clinical trial, the quality and completeness of the data is not perfect (we have described each countries data completeness rates in detail the report). However, we believe the broad epidemiological information presented is of interest and we hope that individual surgeons, hospitals and countries will use the data in the report to further improve the quality of data they collect, and to understand variance in clinical practice and further improve clinical quality.