Is age an independent determinant of mortality in cardiac surgery as suggested by the EuroSCORE?

Background The proportion of older patients in cardiac surgery is continuously increasing. 37% of patients undergoing heart surgery in Germany in the year 2000 were 70 years of age and older. We have studied the role of age as a determinant of mortality in cardiac surgery in our institutional patient population. Methods We have calculated the EuroSCORE and the corresponding age-adjusted EuroSCORE in 8769 patients who underwent heart surgery between January 1996 and January 2002 and collected the information on the occurrence of postoperative complications and 30-days mortality. Results The multimorbidity increased with ascending age. Both the EuroSCORE and the age-adjusted EuroSCORE values increased significantly with age in the whole group of patients as well as in the group of patients who were alive 30 days after heart surgery. The incidence of postoperative complications and 30-days mortality increased significantly with age. In patients who died within 30 days after surgery, the EuroSCORE increased significantly with age, whereas the age-adjusted EuroSCORE did not. The occurrence of diabetes mellitus, arterial hypertension and atrial fibrillation, i.e., the risk factors not considered by the EuroSCORE, exhibited a significant age dependence in our patients. The univariate analysis identified the significant dependence of 30-days mortality on diabetes and atrial fibrillation. The stepwise logistic regression analysis showed the dependence of mortality on diabetes. Conclusions On the background of the well-known age-dependent structural and functional changes of different body organs, our data show that age is a significant risk indicator in cardiac surgery, strongly correlating with morbidity and mortality. Consequently, special preventive and therapeutic measures are required in clinical environment in the case of elderly patients undergoing cardiac surgery.


Backgraound
The proportion of older patients in cardiac surgery is increasing as a consequence of demographic changes and new developments in medical technology. In the year 1989, 3673 patients in the age of 70 years and over underwent a cardiac surgical procedure in Germany, whereas in the year 2000 the number of operated patients in this age group increased to 35884. Thus, the proportion of patients aged 70 years and older increased from 11,2% to 36,7% [1,2].
The EuroSCORE is one of the established risk evaluation scores in heart surgery in Europe [3][4][5]. In Germany, the EuroSCORE is presently being employed as a nation-wide quality control tool for the assessment of the 30-days mortality risk in cardiac surgery. It is based on a logistic regression model and includes all adult cardiac surgery procedures within a defined 30-days time interval. Besides other parameters, the age above 60 years is regarded as an independent determinant of mortality in this score system, whereby the risk weight is increased in 5-years intervals (table 1; see also [http://www.euroscore.org/ calc.html]). We have checked the effect of age on the mortality rate and the occurrence of postoperative complications in our institutional patients.

Methods
Several patient files, for instance, the databases of the departments of anaesthesiology, clinical laboratory and institutional management, the database of the nation-wide quality control in cardiac surgery, internal follow-up information files and the databases of specific groups of patients, e.g. the patients requiring haemodialysis, are assembled in parallel in our institution. In cooperation with the Department of Neuroinformatics at the University in Bielefeld, we have created a new specialized database, bringing together the information from the already existing database systems. Using this specialized database, we have retrospectively studied a total of 8769 patients who underwent a cardiac surgery in the time interval between January 1996 and January 2002.
According to the established EuroSCORE evaluation procedure, our patients were divided into seven age groups. The EuroSCORE and the age-adjusted EuroSCORE values were calculated for each patient, in the latter case by subtracting the scoring points for the corresponding age group. Thereafter, the occurrence of postoperative complications and the 30-days mortality were determined. Finally, statistical analyses (Pearson' Chi square test, ANOVA) were performed with the aim to establish the role of age in cardiac surgery. Statistical significance was defined as p < 0.05. In addition to the EuroSCORE parameters, we have also determined the occurrence of arterial hyperten-sion, diabetes mellitus, as well as atrial fibrillation in all age groups studied. The dependence of the 30-days mortality on these three variables was checked by univariate as well as stepwise logistic regression analyses.

Results
The calculated EuroSCORE parameters in all age groups studied, as well as the occurrence of the three additionally determined accompanying diseases, are shown in table 2. Highly significant age-dependent differences in the distribution of the following variables were found: female gender, chronic obstructive pulmonary disease (COPD), noncardiac atherosclerosis, neurological dysfunction, instable angina, left ventricular ejection fraction 30-50%, non-isolated coronary artery bypass grafting (CABG), thoracic aortic surgery, arterial hypertension, diabetes and atrial fibrillation.   Tables 4 and 5 summarize the surgical procedures performed in the seven age groups studied and show the corresponding 30-days mortalities. The relative number of coronary bypass surgeries, with a mortality rate of 1,1%, decreased with age in our patient population, whereas the number of aortic valve replacement procedures, either with or without accompanying coronary grafting, increased. The mortality for aortic valve replacement was 2,7% and 4,1% in the case of isolated and combined surgical procedures, respectively. Table 6 shows the incidence of postoperative complications. The 30-days mortality and the occurrence of most postoperative complications rose significantly with increasing age. The univariate analysis revealed a significant dependence of 30-days mortality on diabetes (p = 0,016) and atrial fibrillation (p < 0,005). The stepwise logistic regression analysis showed a significant dependence of mortality on diabetes (p = 0,050).

Discussion
The demographic changes result in a continuously increasing number of elderly patients being treated in cardiac surgery departments. In Germany, there exists no nation-wide database on the prevalence and incidence of those accompanying diseases studied in our patient group for the whole population, therefore, it is not possible to compare our patients with an age-matched group of subjects to determine any possible biological selection. If we consider the age-adjusted EuroSCORE as a measure of multimorbidity, then our data show an increase in multimorbidity in patients undergoing cardiac surgery with age. A detailed analysis of the individual EuroSCORE variables and the additional evaluation of other parameters confirm that the number of accompanying diseases in patients undergoing cardiac surgery increases with age. When compared with other age groups, the relatively small number of patients aged 80 years and over in our  [7,8]. If we take this increasing multimorbidity into account, then the rise in 30-days mortality and occur-rence of most postoperative complications with age, observed in our patient population, is not an unexpected result. Moreover, it is in harmony with the results of other research groups [6][7][8][9]. Also the reduction of the number of isolated coronary surgeries with increasing age, observed in our group of patients, corresponds with the findings of other authors [6,8]. This surgical procedure is associated with a lower overall mortality when compared with the valve replacement and repair, as well as combined surgical procedures, nevertheless, it contributes to the increasing mortality with age. In the year 2000, the mortality for isolated CABG in Germany was 2,9%, 3,3% for isolated aortic valve replacement and 6,0% for simultaneous coronary bypass surgery with accompanying aortic valve replacement [2]. This issue is considered in the  EuroSCORE evaluation in the risk factor called "other than isolated CABG".
In the patients, who died within 30 days after surgery, the EuroSCORE rose significantly with age, however, the score value did not show such significant changes after adjustment for age. Therefore, the question arises what is the reason for the increased mortality observed in the elderly and, further, what other factors are concealed behind the common variable 'age'. The following four points have to be discussed with respect to this issue: 1. Several diseases, e.g., COPD, peripheral atherosclerosis or renal insufficiency, are considered in the EuroSCORE analysis, however, the stage of the respective diseases and the intensity of the pathological processes, both of which are influenced by the time factor, are not taken into account. Therefore, the different intensity and duration of the mentioned diseases may be hidden in the EuroSCORE parameter 'age'.
2. The EuroSCORE risk evaluation system has been derived from a large database of cardiac surgical patients. Some diseases were not identified as being relevant with respect to mortality in this large patient population by the multivariate analysis and, therefore, were not considered during the generation of the EuroSCORE list of risk factors. We have studied the dependence of mortality on arterial hypertension, diabetes and atrial fibrillation, i.e., the three factors not included in the EuroSCORE system. All three factors exhibited significant differences in their respective age-related distributions. The univariate analysis determined a significant dependence of the 30-days mortality on diabetes and atrial fibrillation. The multivariate analysis revealed the dependence of the 30-days mortality on diabetes. The effect of such factors, that are not separately considered in the EuroSCORE analysis, may also partly contribute to the parameter 'age'.
3. The following factors are taken into account by the Eu-roSCORE system to characterize the cardiac status: instable angina, left ventricular ejection fraction 30-50% or <30%, myocardial infarction within preceding 90 days before surgery, systolic pulmonary artery pressure >60 mm-Hg. These patient data guarantee objective data acquisition and good comparability among hospitals with differing personnel and technology backgrounds and adequately describe certain aspects of the cardiac status. It would be inappropriate to expect the EuroSCORE system to consider in detail all characteristic features of the cardiovascular system in its complexity. However, one factor, shown to possess a high degree of diagnostic and therapeutic relevance in the elderly, the diastolic cardiac function [10], is not taken into account in the EuroSCORE cardiac operative risk evaluation. Senni et al. [11] have found out that 43% of patients with the primary diagnosis of cardiac failure exhibited normal systolic heart function.
Other studies have shown that diastolic heart failure is found in about 50% of elderly patients suffering from congestive heart failure [12,13]. This means, that also this aspect may be hidden in the EuroSCORE parameter 'age'. 4. Finally, the aging process is associated with structural and functional changes in various organ systems, which may influence the perioperative outcome. The following age-dependent changes in the cardiovascular system have previously been identified: dilation of large vessels, thickening of vessel walls, affecting first of all the intimal layer, loss of vessel elasticity [14][15][16], increase in left ventricular afterload [17], thickening of left ventricular wall [14,18], cardiac myocyte cell loss and increase in the amount of myocardial collagen [19], decline in early diastolic left ventricular filling rate [20], increased atrial contribution to ventricular filling and larger atrial diameter [18], reduced adaptability of the cardiac response to different workloads resulting from modified ventricular filling volumes and changes in heart rate [17], increasing deficits in sympathetic modulation [21], changes in myocardial calcium hemostasis, affecting the excitation-contraction coupling and, consequently, the myocardial contractility and tendency to arrhythmias [22,23]. The consequence of all the above-mentioned changes is the reduced cardiac adaptability to workload observed in the elderly [24,25].
The age-dependent changes in other than cardiovascular body organ systems require specific preventive and therapeutic measures. For instance, the well-known age-related changes in lung structure and function (e.g., increase in functional residual capacity and residual volume, weakness of the expiratory muscles, malfunction of the airway epithelium accompanied with increased production of mucus, diminished tendency to cough, increased ventilation-perfusion mismatch) may lead to complications in older patients undergoing cardiac surgery. To prevent pulmonary complications in the elderly, it is necessary to examine the lung function and to initiate chest physiotherapy with breathing exercise already preoperatively. Early extubation, mobilisation, and pharmacologically supported bronchial clearance should follow postoperatively and a dehydration of the patients should be avoided. (A review of other therapeutic measures in elderly cardiac surgical patients is given in [26].)

Conclusions
It can be summarized that age strongly correlates with a cluster of risk factors and organ dysfunctions, occurring during aging. However, the old age as such is not a disease. It is only a state characterized by a relatively high probability of suffering from various disturbances of the normal body function. From this point of view, age is a risk factor strongly correlating with mortality. Because of increased occurrence of age-related multimorbidity and changes in organ structure and function, special attention has to be paid to adequate therapeutic measures in elderly patients undergoing heart surgery.