Determinants of Outcome in Adult Cardiac Surgery

Determinants of Outcome in Adult Cardiac Surgery
Acta Universitatis Tamperensis, No. 1554

By Juha Nissinen
November 2010
Tampere University Press
Distributed by Coronet Books
ISBN: 9789514482311
216 pages
$79.50 Paper Original

A total of 4,563 adult patients underwent cardiac surgical procedures at Vaasa Central Hospital from January 1994 to June 2008. A substantial amount of data from all the patients was prospectively collected in an institutional database, including both cardiac examinations and surgical care. This material was later retrospectively analysed in five studies. The main focus of these studies was on recognizing specific risk factors related to various cardiac surgical procedures. The first two studies evaluated the risk predicting value of preoperative spirometry.

In the first one we showed that impaired left ventricular function (EF<50%) is strongly associated with preoperative low percentages of forced vital capacity (FVC<80%) in a material of 453 patients undergoing aortic valve reconstruction (AVR). The same association was not found with forced expiratory volume in 1 second (FEV1). Preoperatively lowered FVC (FVC<80%) had a better prognostic value for postoperative death in AVR patients than preoperatively lowered FEV1 (FEV1<75%). Also, the lowered FVC had a clear predictive value for adverse neurologic events and prolonged postoperative stay in the intensive care unit (ICU). The second study ascertained the prognostic value of preoperative spirometry among 1,848 patients undergoing coronary bypass surgery (CABG). The percentages of predicted FVC and of predicted FEV1 were associated with in-hospital death, combined adverse end-point, need for postoperative de novo dialysis, neuropsychological disturbances, atrial fibrillation as well as length of stay in the intensive care unit ? 5 days. Only percentage of predicted FVC < 70% along with pulmonary disease but not percentage of predicted FEV1 < 70%, were independent predictors of late overall mortality.

The prevalence of people aged 80 years or over is steadily increasing. The third study included a consecutive series of 247 patients aged 80 years or over who 8 Determinants of Outcome in Adult Cardiac Surgery underwent isolated CABG. They belonged to a series of 3,474 patients who underwent isolated CABG during the same study period. Of them 40% were women (24% among the younger age group). A special focus was on the long-term survival (5 years) of these patients and on risk factors contributing to it. Both univariate analysis and regression analysis showed that diabetes, extracardiac arteriosclerotic disease, neurologic dysfunction, recent myocardial infarction and critical preoperative status were associated with poorer long-term outcome. The 30-day in-hospital mortality in this material was 4.7% compared to 1.3% in the younger age group. The 5-year survival rate was 77% and the 10-year survival 35% (compared to 90% and 76% in the younger group). A propensity score analysis between matched pairs of patients aged 80 years or over and under 80 years showed parallel survival curves between these two groups up to five years after the greater initial 30-day mortality of older patients.

There is longstanding evidence confirming that a long aortic cross-clamp time (XCT) and a long cardiopulmonary bypass time (CPBT) are strong risk factors connected to heart surgery. In our fourth study both XCT and CPBT were cut in cohorts of 30-minute intervals. Both XCT and CPBT were included separately into a statistical regression model in proportion to changes in mortality rate.

The best cutoff values of increasing risk were 150 min for XCT and 240 min for CPBT (adjusted for additive EuroSCORE and complexity of the operation). It is worth noting that these cutoff values, despite having a high accuracy (over 90%), have a very low sensitivity (34% for XCT and 28% for CPBT). Thus long XCT and CPBT do not automatically indicate a dismal prognosis. XCT and CPBT have, however, a marked impact on postoperative morbidity. The strong association between postoperative stroke and XCT as well as CPBT is of particular interest. In this series, CPBT was a much stronger predictor of 30-day mortality than XCT.

EuroSCORE, established in 1999, is the most popular and widely validated risk-scoring system in Europe. It is known that EuroSCORE tends to overestimate mortality in low-risk patient groups and to underestimate the risk in very highrisk patient groups. In the original EuroSCORE, non-CABG surgery was calculated as a risk. Thus a simple AVR carries greater risk than a CABG with several bypasses. On the other hand, the complexity of the operation and the severity of the non-cardiac illnesses of the patient are not fully evaluated. In the fifth study we created a new model for risk evaluation (Modified EuroSCORE). More attention was paid to preoperatively impaired renal function by estimating the glomerular filtration rate. Also, the complexity of the operation was taken into consideration.

Not surprisingly, this model fitted our material well in Vaasa. The model was tested and validated in Tampere University Hospital (study VI), with an adult cardiac surgery material of 4,014 people, operated from January 2004 to December 2008.

Here, the predicted mortality rates with the Modified EuroSCORE were rather similar to those observed. This was not the case for EuroSCORE. Interestingly, the accuracy of the modified score was particularly evident in high risk patients.

Return to Coronet Books main page