Ontario score and cardiac risk during waiting for elective coronary bypass grafting

International Journal of Cardiology 110 (2006) 167 – 174 www.elsevier.com/locate/ijcard

Ontario score and cardiac risk during waiting for elective coronary bypass grafting Fernando Henpin Yue Cesena, Desiderio Favarato, Luiz Antoˆnio Machado Ce´sar Se´rgio Almeida de Oliveira, Prota´sio Lemos da Luz * Heart Institute (InCor), University of Sa˜o Paulo Medical School, Av. Dr. Ene´as de Carvalho Aguiar 44, SP 05403-000 Sa˜o Paulo, Brazil Received 13 February 2005; received in revised form 19 June 2005; accepted 26 June 2005 Available online 1 August 2005

Abstract Background: Waiting lists for coronary bypass grafting are of major concern in several countries and prioritisation systems to the surgery have been proposed. The aim of this study was to verify the adequacy of Ontario score in predicting cardiac events during the waiting for elective coronary bypass grafting. Methods: A composite end-point (sudden or cardiac death, myocardial infarction, unstable angina or hospital admission) and sudden, cardiac death were analysed in 460 patients referred to the surgery. The relation between Ontario score and events was verified. Results: Median waiting time was 126 days. The composite end-point and sudden, cardiac death occurred in 21.7% and 2.7% of the cases, respectively. In relation to Ontario score  6.00, considered the lower-risk subset, only patients in score <4.00 (7.2% of whole study population) presented a higher chance of the composite end-point during the waiting. ROC curve did not show adequate accuracy of Ontario score in predicting the composite end-point (area under the curve 0.53, p = 0.36). Ontario score could not predict the risk of death. Total complications and death occurred within acceptable waiting times by Ontario recommendation in 47.8% and 36.4% of the cases, respectively. Waiting longer than maximum wait defined by Ontario was not associated with an excess of complications. Conclusions: Ontario score showed a limited value in predicting cardiac events during the waiting for elective coronary bypass grafting. The results emphasise the need for shortening the wait in order to reduce complications in the period. D 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: Coronary disease; Coronary artery bypass; Waiting list; Complications; Risk assessment

1. Introduction Waiting lists for coronary artery bypass graft surgery are frequent in countries that provide universal access to healthcare. Avoidable cardiac complications, hospital admissions and costs, deterioration of left ventricular function, mental stress and impairment of quality of life are undesirable consequences of delayed waiting for surgery [1– 6]. Therefore, deciding who deserves early operation and who can wait longer assumes great importance. Priority criteria have been proposed; one of these, the Ontario score

* Corresponding author. Tel.: +55 11 30695352; fax: +55 11 30695447. E-mail address: [email protected] (P.L. da Luz). 0167-5273/$ - see front matter D 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijcard.2005.06.047

[7], was elaborated by a panel consensus through analyses of 438 fictitious cases and considers classical prognostic factors in coronary artery disease: anginal status, coronary anatomy, ischemia at non-invasive tests and left ventricular function (Appendix 1). Ontario score establishes maximum acceptable waiting times according to the urgency for surgery [7]. Morgan et al. [8] reported that waiting for the surgery longer than recommended by Ontario score is an independent predictor of mortality during the waiting period. On the other hand, studies from New Zealand [9,10] showed limitations of scoring systems in predicting cardiac events. Moreover, a high proportion of complications occur early or within acceptable waiting times in several reports [1 –3,8,11,12], also raising concerns about the usefulness of urgency scales.

168

F.H.Y. Cesena et al. / International Journal of Cardiology 110 (2006) 167 – 174

Accordingly, this study was designed to assess the adequacy of the Ontario score in predicting adverse events in real patients waiting for elective coronary bypass grafting in an out-of-hospital setting.

2. Material and methods From January 1, 1998 to July 12, 2001, 596 patients were eligible for elective coronary bypass grafting at Heart Institute (InCor), Hospital das Clı´nicas, University of Sa˜o Paulo Medical School. Patients affected by valve disease that required surgical correction were not included, but those amenable to left ventricular reconstruction combined to coronary bypass graft surgery were accepted. Excluding 22 cases in which therapeutic choice was reconsidered and crossed over to coronary angioplasty/stenting or medical treatment, 58 cases of reoperation (for which complex anatomic changes make difficult precise determination of Ontario score) and 56 patients in whom the score could not be calculated due to missing data, the study population comprised 460 individuals. Patients were followed in the setting of a specialised Coronary Artery Disease Outpatient Clinic. The authors did not interfere in the usual selection criteria used to schedule surgeries, which included anginal status, coronary anatomy, ischemia at non-invasive tests and left ventricular function. The study was approved by the institutional Ethics Committee. Data were collected by a single researcher (FHYC) through interviews and analysis of medical records. Time for surgery or complication was always considered from the moment of referral to surgery. Separate analyses were performed for three types of events: (1) a composite endpoint, which included sudden or cardiac death, myocardial infarction, unstable angina or hospital admission due to cardiac cause for at least 24 h; (2) the need for urgent surgery, defined as that procedure that was advanced, occurring in the same hospitalisation of a non-fatal complication; and (3) sudden or cardiac death. Ontario score was calculated at the moment of referral to surgery as described by Naylor et al. [7] (Appendix 1). For determination of angina grade, the modified Canadian Cardiovascular Society classification [7] was considered (Appendix 2). High-risk ischemia at non-invasive tests was considered if at least one of the following features was present: & at exercise ECG testing: ST depression higher than 0.2 mV if upsloping or equal to or higher than 0.2 mV if horizontal or downsloping, ST depression of at least 0.1 mV in the first stage or persistent ST depression after 5 min of recovery; & at perfusion scintigraphy: multiple reversible perfusion defects, extensive reversible perfusion defect, transient left ventricle dilation after stress or increased lung thallium-201 uptake after stress;

& at stress echocardiogram: transient motility alterations suggestive of ischemia in multiple walls, corresponding to more than one vascular supply region, extensive transient motility alteration and reversible left ventricle dilation after stress. Left ventricular systolic function was primarily assessed by ventriculography during cardiac catheterisation before referral to surgery. The function was graded according to the Coronary Artery Surgery Study (CASS) [13]. Normal function was considered if the score was 5 or 6; mild dysfunction if the score summed 7 to 9, moderate if 10 to 12 and severe if above 12. In the absence of ventriculogram, left ventricular systolic function was determined by transthoracic echocardiogram and categorised into normal and mild, moderate or severe dysfunction. Coronary lesions were classified as proposed by CASS [13]. Maximum acceptable time for surgery, according to Ontario score, was also described by Naylor et al. [7]: score 1 (emergency): immediate surgery; 2 (extremely urgent): surgery within 24 h; 3 (urgent): 24 to 72 h; 4 (semi-urgent): 72 h to 14 days; 5 (short list): 2 to 6 weeks; 6 (delayed): 6 weeks to 3 months; and 7 (marked delay): 3 to 6 months. 2.1. Statistical analyses Kaplan– Meier product limit method was performed to establish the relation between Ontario score and complications. For the composite end-point, Ontario score was categorised into four groups: <4.00, 4.00 –4.99, 5.00– 5.99 and  6.00. For fatal events, different analyses were performed with Ontario score divided into two categories (cut-off points 4.50, 5.00, 5.50 and 6.00). Kaplan –Meier method was also used to evaluate the risk of events according to two groups defined whether the waiting time exceeded or not that recommended by Ontario. In all these analyses, the time for the first event was considered. For censored cases, that is, when no complication was observed, the time considered was the waiting time for surgery, the time until the patient was withdrawn from the waiting list due to any cause, or the time between referral to surgery and the last clinical evaluation (when the patient was waiting for the surgery at February 28, 2002, when the authors stopped collecting data). Comparisons among Kaplan– Meier curves were performed by log-rank test. Hazard ratios (HR) for complications (and respective confidence intervals [CI]) were calculated by Cox regression. The same method was used to determine independent variables related to the need for urgent surgery. ROC (receiver operating characteristic) curve was also analysed to verify accuracy of Ontario score in predicting the composite end-point. Waiting times and Ontario scores were compared by Kruskal Wallis and Mann – Whitney tests. Student’s t and chi-square tests were performed to compare other continu-

F.H.Y. Cesena et al. / International Journal of Cardiology 110 (2006) 167 – 174

169

Table 1 Baseline characteristics and laboratory data during the waiting for the surgery of total study population, groups ‘‘in’’ and ‘‘out’’ (whether or not the surgery occurred within time recommended by Ontario, respectively), and groups ‘‘urgent’’ and ‘‘elective’’ (if the surgery was anticipated occurring in the same hospitalisation of a complication or it was elective, respectively) Variablea

Age (years) (457) Male gender (%) (460) Arterial hypertension (%) (460) Diabetes mellitus (%) (458) Obesity (%) (436) Total cholesterol (mg/dl) (409) LDL-cholesterol (mg/dl) (247) HDL-cholesterol (mg/dl) (254) Triglycerides (mg/dl) (407) Angina functional class III – IV (%) (458) Severe left ventricular dysfunctionb (%) (438) 3-vessel disease (%) (459) High-risk ischemia at non-invasive test (%) (179) Ontario score (%) (460)

Ontario score (460) Waiting time (days) (396)

Total (n = 460)

<4.00 4.00 – 4.99 5.00 – 5.99  6.00

Groups ‘‘in’’ (n = 144)

‘‘out’’ (n = 284)

60.8 T 10.0 72.8 78.5 37.3 22.2 215 T 45 139 T 40 37 T 9 180 T 101 27.7

60.4 T 9.4 75.7 74.3 37.5 18.8 215 T 47 131 T 34 37 T 10 191 T 133 28.9

61.0 T 10.2 72.9 81.3 38.7 23.3 214 T 43 141 T 41 38 T 9 178 T 111 28.5

22.4

18.7

56.4 31.3 7.2 6.1 30.2 56.5 5.89 T 0.89 170 T 157

p ‘‘in’’ vs. ‘‘out’’

Groups

p ‘‘urgent’’ vs. ‘‘elective’’

‘‘urgent’’ (n = 53)

‘‘elective’’ (n = 345)

0.42 0.53 0.09 0.78 0.30 0.73 0.07 0.55 0.35 0.94

63.1 T 9.6 66.0 88.7 37.7 20.8 214 T 44 135 T 33 32 T 9 217 T 106 47.2

60.5 T 10.0 74.5 77.4 39.0 21.3 214 T 46 138 T 40 38 T 9 180 T 125 26.5

25.3

0.13

13.7

23.3

0.13

50.7 23.5

61.5 36.2

0.03 0.11

60.4 36.4

56.7 30.1

0.61 0.67

1.4 3.5 20.8 74.3 6.28 T 0.65 63 T 40

10.6 7.4 34.1 47.9 5.69 T 0.94 227 T 166

15.1 9.4 24.5 50.9 5.63 T 1.10 90 T 90

6.7 5.5 29.0 58.8 5.94 T 0.87 183 T 161

0.10

<0.01

<0.01 <0.01

0.08 0.19 0.06 0.87 0.98 0.94 0.70 <0.01 0.04 <0.01

0.05 <0.01

Data are expressed as percentages (calculated excluding missing data) or mean T standard deviation. a Number in parentheses refers to patients in whom the information was available. Laboratory, myocardial ischemia and left ventricular function data refer to assessment during the waiting period. For a single patient, mean value of a variable was considered if it was assessed more than once during the waiting. b By ventriculography during cardiac catheterisation or echocardiogram.

ous variables and proportions among categorised variables, respectively. When appropriated, Ontario score was rounded to a whole number. In all analyses, a p value lower than 0.05 was considered significant. Data management and analyses were performed using SPSS for Windows version 10.0 (SPSS Inc., Chicago, Illinois, USA).

3. Results 3.1. Patient characteristics, waiting times and follow-up Baseline characteristics and laboratory data during the waiting period are given in Table 1. Ontario score ranged from 2.95 to 7.15, the mean T SD was 5.89 T 0.89 and the median was 6.15. Mean waiting time for surgery was 170 T 157 and the median time was 126 days. The higher the score, the longer was the waiting time: 121 T119, 145 T 132, 170 T 159 and 180 T 162 days for scores < 4.00, 4.00 – 4.99, 5.00 – 5.99 and  6.00, respectively ( p = 0.05). Among the 460 patients enrolled in the study, the great majority (404 or 87.8%) had been subjected to surgery at the end of the study, 12 (2.6%) died while waiting, one (0.2%)

was withdrawn from the waiting list after stroke, 23 (5.0%) refused operation, three (0.7%) were still awaiting surgery when data collection was stopped and in 17 (3.7%) the follow-up information could not be obtained. 3.2. Ontario score and the composite end-point The composite end-point during the waiting was observed in 21.7% of available data (93 / 429). Free probability of the composite end-point during the waiting, according to Ontario score categories, is shown in Fig. 1. An overlap can be noted among curves for scores  4.00 and the observed significance level can be solely attributed to a higher chance of events in score <4.00, which represented only 7.2% of the whole study population. In fact, in relation to score  6.00, only score <4.00 showed significantly increased risk of the composite end-point according to multivariate Cox regression analysis adjusted to gender, age and main risk factors (LDL-cholesterol, HDL-cholesterol, triglycerides, arterial hypertension, diabetes mellitus and obesity) (Table 2). The influence of the factors used in the Ontario score composition was also assessed. The results did not change and only Ontario score <4.00 persisted significantly related to complications after the addition of left ventricular

F.H.Y. Cesena et al. / International Journal of Cardiology 110 (2006) 167 – 174 1.00

1.0 0.9

4.00-4.99

0.8 0.7 5.00-5.99

0.6

0.75

≥ 6.00

0.5 < 4.00

0.4

sensibility

free probability of the composite end-point

170

0.3 p = 0.01

0.2

0.50

0.1 0

90

180

270 360 450 waiting time (days)

540

A = 0.53 (0.46-0.60) p = 0.36

0.25

630

Fig. 1. Free probability of the composite end-point during the waiting for coronary bypass graft surgery, according to Ontario score categories. 0.00

function (presence or absence of severe dysfunction) or coronary anatomy (according to categories of Ontario score [Appendix 1]) to the model described above (HR = 6.6 and 4.8, p < 0.01 for Ontario score < 4.00 vs.  6.00, respectively). However, Ontario score < 4.00 lost its predictive value after the model was adjusted to anginal status (functional class III – IV or I– II) or ischemia at non-invasive tests (presence or absence of high-risk ischemia). Analysis of ROC curve showed a non-significant accuracy of Ontario score in predicting the composite end-point (Fig. 2). 3.3. Ontario score and the need for urgent surgery Among patients presenting a non-fatal composite endpoint, in 53 the surgery was anticipated occurring in the same hospitalisation of the complication. This ‘‘urgent’’ group was compared to 345 individuals in whom the surgery was elective (Table 1). At baseline, patients in the ‘‘urgent’’ group were characterised by worse anginal status, lower HDL-cholesterol, higher triglyceride levels and a trend for older age and higher proportion of arterial hypertension. The relation between Ontario score and the need for urgent surgery, at multivariate Cox regression, resembled what was observed to the composite end-point. In relation to Ontario score  6.00, score < 4.00 was predictive of need for urgent surgery ( p < 0.01); however, the

0.50 1- specificity

0.25

0.75

1.00

Fig. 2. ROC curve for the ability of Ontario score in predicting the composite end-point during the waiting for coronary bypass graft surgery. A=area under the ROC curve (95% confidence interval).

significance of this association diminished when Ontario scores 4.00 – 4.99 and 5.00 – 5.99 were compared to  6.00 ( p = 0.05 and 0.82, respectively). Importantly, the relation of Ontario score and urgent surgery was largely due to angina functional class. Furthermore, low HDL-cholesterol and high triglycerides levels remained significantly related to urgent surgery in this multivariate analysis ( p < 0.01 and = 0.02, respectively). 3.4. Ontario score and sudden or cardiac death Sudden or cardiac death occurred in 2.7% of available complete data (12/442), specifically one in score 4, one in score 5, five in score 6 and five in score 7. As shown in Table 3, Ontario score, categorised at different thresholds, could not predict the risk of fatal complications. 3.5. Time effects and the Ontario score recommendation The higher the Ontario score, the higher was the estimated probability of surgery occurring within time recommended by Ontario (Fig. 3). The proportion of welltimed operation was 0.0% (0 / 5), 9.5% (4 / 42), 12.5% (7 / 56), 35.2% (70 / 199) and 61.7% (58 / 94) in scores 3, 4, 5, 6

Table 2 Hazard ratios for the composite end-point, according to Ontario score categories, in relation to score  6.00 Ontario score

HR

95% CI

p

<4.00 4.00 – 4.99 5.00 – 5.99

4.8 1.7 1.1

1.8 – 12.7 0.5 – 5.9 0.5 – 2.3

<0.01 0.37 0.82

CI=confidence interval and HR=hazard ratio. Cox regression analysis adjusted to gender, age (1-year increment), lipid levels (LDL-cholesterol, HDL-cholesterol, triglycerides, all 1-mg/dl increment), presence or absence of arterial hypertension, diabetes mellitus and obesity.

Table 3 Hazard ratios for sudden or cardiac death according to Ontario score categorised at different thresholds Comparison- scores <4.50 <5.00 <5.50 <6.00

vs. vs. vs. vs.

4.50 5.00 5.50 6.00

HR

95% CI

p

1.6 1.1 0.7 0.5

0.2 – 12.9 0.1 – 9.3 0.1 – 3.2 0.1 – 2.0

0.67 0.90 0.63 0.37

CI=confidence interval and HR=hazard ratio.

F.H.Y. Cesena et al. / International Journal of Cardiology 110 (2006) 167 – 174 1.0

4. Discussion

0.6 0.4 0.2 0.0 90

180

270 360 450 waiting time (days)

540

630

Ontario < 4.00

Ontario 4.00-4.99

Ontario 5.00-5.99

Ontario ≥ 6.00

Fig. 3. Probability of surgery within time recommended by Ontario, according to score categories.

and 7, respectively ( p < 0.01). Overall, this proportion was 35.1% (139 / 396, available data). Mean time for occurring the composite end-point was 114 T 127 and the median 65 days. The proportion of patients in whom complications occurred within acceptable waiting times by Ontario recommendation was 0.0% (0 / 2), 30.8% (4 / 13), 15.4% (2 / 13), 52.6% (20 / 38) and 70.8% (17 / 24) in scores 3, 4, 5, 6 and 7, respectively ( p = 0.01). Including all patients, this proportion was 47.8% (43 / 90, available data). Time to death was available in 11 patients, the mean was 247 T 221 and the median 222 days. Overall, four in 11 patients (36.4%) died within maximum wait by Ontario. To further characterise the adequacy of Ontario score, the hypothesis that waiting more than recommended by Ontario would be associated with a higher chance of complications was tested. Two groups of patients were compared, named groups ‘‘out’’ (n = 284) and ‘‘in’’ (n = 144), defined if the waiting time surpassed that recommended by Ontario or not, respectively. As shown in Table 1, no higher risk features were identifiable in baseline characteristics of group ‘‘in’’. On the contrary, group ‘‘out’’ showed a lower Ontario score, a significant higher proportion of patients with three-vessel disease and a trend for more prevalent arterial hypertension and higher LDL-cholesterol levels. Despite these facts, patients in group ‘‘in’’ had a significantly higher risk of the composite end-point and sudden, cardiac death along the time (Fig. 4). The rate of the composite end-point was 23.9% and 19.7% in groups ‘‘in’’ and ‘‘out’’, respectively. Fatal event occurred in 2.8% in group ‘‘in’’ and in 2.5% in group ‘‘out’’. Importantly, group ‘‘in’’ concentrated individuals whose complications motivated an anticipation of the surgery, allowing it to occur within time recommended by Ontario. Considering only patients who presented the composite end-point, the proportion of individuals in whom the event advanced the operation, occurring in the same hospitalisation, was 86.7% in group ‘‘in’’ and 55.1% in group ‘‘out’’ ( p < 0.01).

The main findings of this study can be summarised as follows: (1) patients with Ontario score < 4.00 (only 7.2% of whole study population) presented a higher chance of complications and need for urgent surgery during the waiting for coronary bypass grafting; for most patients, that is, those with scores  4.00, no ability of Ontario score in predicting the composite end-point was detected; (2) Ontario score was not able to predict fatal complications; (3) a high proportion of events occurred within acceptable waiting times by Ontario recommendation; and (4) it was not possible to demonstrate a higher chance of complications associated with waiting more than that recommended by Ontario. Actually, the group subjected to the surgery within the time recommended by Ontario was associated with a higher chance of events. This finding may be

(a) 1.0 0.9

free probability of the composite end-point

0.8

Group “out”

0.8 0.7 0.6 Group “in” 0.5 0.4 p < 0.01 0.3 0.2 0.1 0.0 0

90

180

270

360

450

540

630

waiting time (days)

(b) Group “out”

1.0 free probability of sudden, cardiac death

estimated proportion of well timed surgery

p < 0.01

0

171

0.9

Group “in”

0.8 0.7

p = 0.01

0.6 0.5 0.4 0.3 0.2 0.1 0.0 0

90

180

270 360 waiting time (days)

450

540

630

Fig. 4. Free probability of the composite end-point (a) and sudden or cardiac death (b) during the waiting for coronary bypass graft surgery, according to groups ‘‘in’’ (waiting time within that recommended by Ontario) and ‘‘out’’ (waiting time exceeded that recommended by Ontario).

172

F.H.Y. Cesena et al. / International Journal of Cardiology 110 (2006) 167 – 174

interpreted as a consequence of group ‘‘in’’ that have concentrated those individuals who were operated in an urgent basis after an acute coronary event, together with the relatively high frequency of complications in the population study. The use of risk scores to predict cardiovascular and coronary events has been widely advocated to guide preventive and therapeutic decisions [14 – 16]. In the setting of waiting for coronary bypass grafting, several cardiac surgery centres prioritise patients according to predictive factors or urgency scales [1– 3,8,17]. We chose to assess the accuracy of the Ontario score as representative of what is done in most centres, including ours, once it considers the classical risk factors in coronary artery disease. In fact, our results showed that the higher the Ontario score, the longer was the waiting time. Moreover, Ontario score is relatively simple, objective and easy to calculate, allowing comparisons among different reports. The limited value of Ontario score in predicting events described in this study agrees with reports by Jackson et al. [9] and Seddon et al. [10], in which waiting times were even longer than ours (median 212 and 146 days to elective surgery, respectively). In contrast, priority status and waiting more than that recommended by Ontario have been shown to be independently related to mortality during the waiting in reports by Rexius et al. [17] and Morgan et al. [8], respectively. These last studies have major differences to ours regarding patient selection and waiting times, making difficult direct comparisons. Our report included only elective patients and the median waiting time was 126 days, sharply contrasting to 55 and 18 days in the publications by Rexius et al. and Morgan et al., respectively, in analyses that comprised both elective and urgent cases. Indeed, while only 35% of our sample underwent surgery within acceptable time by Ontario, this proportion was 75% in Morgan’s study. Regarding temporal distribution of events, our results compare with several other reports which showed a high proportion of complications early in the waiting or within acceptable times according to risk models [1 –3,11,12]. Even in the Ontario’s experience, 66% of deaths happened within recommended maximum wait [8]. Therefore, our results join to other evidences that raise concerns about the adequacy of risk scales and specially the establishment of ‘‘acceptable’’ waiting times, in the setting of waiting for coronary bypass graft surgery. Indeed, with a possible exception for Ontario’s experience, the real utility of these scores in preventing events has not been convincingly proved in the clinical practice. The reasons for the limited power of these scores must be investigated. One possible explanation is that, for these high-risk patients, the unpredictable feature of coronary disease clinical course is prominent enough to weaken the power of urgency scales. In addition, mortality during the waiting must be analysed under the perspective that a proportion of these patients is of very high risk and will

have a fatal event independently of the therapeutic choice. Deaths during the waiting must be confronted with the existence of operative and early post-operative mortality. Another possibility is that currently used scores are not suitable for predicting events under these circumstances, due to a misuse of predictive factors that compose them. It is very difficult to identify and give the exact importance to the real predictive factors of poor outcome during the waiting period. Ontario score [7] was derived from subjective opinion of specialists who analysed fictitious cases, which can be criticised. It is conceivable that scores developed with predictive factors of complications obtained from real waiting list studies would work better. However, patients enrolled in these studies have already been submitted to some prioritisation system and biases can occur when analyses are performed. For instance, a potential predictor may not appear as a determinant of complication because physicians prioritise patients with this alteration, reducing their waiting time and consequently the likelihood of events. Even so, it is possible that some classical prognostic factors, commonly considered in the formulation of scores, are not predictors of complications in the specific setting of waiting for coronary bypass grafting or their relative weights are different from those considered by the scores. For example, it is assumed that angina status contributes with about 60% of Ontario score, coronary anatomy with 25%, ischemia at non-invasive tests with 15% and left ventricular function with only less than 10% [18]. In contrast, left ventricular dysfunction was shown to be an important and independent predictor of death in the two largest studies on this subject [8,17] and of cardiac events in another [11]. In our previous report that included the population of this study [12], the only independent factors related to total complications were angina, heart failure functional classes and high triglyceride levels. Neither coronary anatomy nor ischemia at non-invasive tests was associated with the composite endpoint and left ventricular dysfunction was related to fatal events, at univariate analysis, but not to the composite end-point. Furthermore, it is possible that important predictive factors are not being considered in the composition of scores. That is the case, for instance, of age, male gender and concomitant aortic valve disease, shown to be independently determinants of death on the waiting list in large studies [8,17,19]. The influence of other risk factors for atherosclerosis may also play a role in the hazard during the waiting period. In this regard, it is noteworthy that our findings are relating features of metabolic syndrome, such as hypertriglyceridemia and low HDL-cholesterol levels, with complications during the waiting [12] and need for urgent surgery. In fact, it must be pointed out that Ontario score was introduced several years ago and, since then, important advances have occurred not only in pathophysiological and epidemio-

F.H.Y. Cesena et al. / International Journal of Cardiology 110 (2006) 167 – 174

logical fields, but also in pharmacological, interventional treatment and coronary surgery indication strategy. As a consequence, risk assessment and prioritisation systems must contemplate these dynamic changes in coronary disease knowledge. Finally, it must be considered that factors implicated with non-fatal complications are not exactly the same of those related to death [12]. On the other hand, non-fatal complications are far more frequent than deaths during the waiting. It can be debated upon and this involves important ethical issues, which is the more appropriate or more costeffective strategy: prioritise a score to prevent the minority but fatal events or, alternatively, give preference to a score that predicts most but non-fatal complications. 4.1. Limitations The small number of deaths is certainly a limitation of this study; however, no trend of association between Ontario score and mortality was verified. The study recruited only patients considered stable enough to await surgery out of hospital. As we verified that those with Ontario score <4.00 had a higher chance of complications, it is possible that the score would be of higher value in subsets of more severely diseased patients. In addition, regional differences in population characteristics and waiting times must be taken into account. As different predictive factors are detected in the analyses of waiting lists from diverse countries, it is expected that the relation between urgency scores and events also vary on a regional basis, limiting extrapolation of our results to other populations. Finally, despite analysing the relation between Ontario score and complications, the study does not provide means for improving accuracy of algorithms. 4.2. Implications and conclusions This study allows the conclusion that Ontario score is of limited value in predicting cardiac events during the waiting for elective coronary bypass grafting, in presence of long delay for the surgery. Despite identifying a subgroup of higher risk (score <4.00), this stratification method did not show satisfactory accuracy in predicting events for most population. The results highlight the difficulty to prioritise these patients based on urgency scales and reinforce the need for improving healthcare provision and shortening the waiting, as delay is highly correlated to mortality before scheduled surgery [12,17]. Using urgency scores and establishing acceptable waiting times must be done with caution and criticism. Furthermore, without undervaluing the importance of classical risk factors, it is possible that consideration of other features more related to patient vulnerability, such as inflammatory factors and plaque composition, may also lead to a more refined risk stratification, optimising prioritisation process and ultimately reducing morbidity and mortality.

173

Appendix A. Ontario score Coronary anatomy

Left main stenosis Multivessel CAD with proximal LAD 3-vessel CAD without proximal LAD 1-vessel CAD with proximal LAD 1 or 2-vessel CAD without proximal LAD Number to be subtracted if high ischemic risk at non-invasive test Number to be added if normal LV function Number to be subtracted if moderate or severe LV dysfunction

Stable angina

Unstable angina

I – II

III

IV-A IV-B IV-C

5.40 6.15 6.45 6.80 6.95

4.85 6.00 6.35 6.55 6.65

4.75 5.50 5.80 5.80 6.15

0.90

0.75

0.75 n/a

0.20

0.20

0.20 0.20 0.20

0.20

0.20

0.20 0.20 0.20

3.40 3.90 3.90 4.05 4.15

2.15 2.55 2.65 2.90 3.05 n/a

CAD=coronary artery disease, LAD=left anterior descending artery, LV=left ventricle, and n/a=not applicable. Modified from Naylor CD et al., Lancet 1990; 335: 1070-1073.

Appendix B. The modified Canadian Cardiovascular Society classification for angina grading Class

Clinical picture

I – II III IV-A

Mild to moderate stable angina on reasonable medical therapy Severe stable angina on reasonable medical therapy Unstable angina; pain resolved with intensified medical therapy, now stable on oral medication Unstable angina on oral medication, symptoms improved but angina with minimal provocation persists Symptoms not manageable on oral medication, requiring coronary care monitoring and parenteral medication; may be hemodynamically unstable

IV-B IV-C

Reference: Naylor CD et al., Lancet 1990; 335: 1070-1073.

References [1] Koomen EM, Hutten BA, Kelder JC, Redekop WK, Tijssen JG, Kingma JH. Morbidity and mortality in patients waiting for coronary artery bypass surgery. Eur J Cardiothorac Surg 2001;19:260 – 5. [2] Silber S, Muhling H, Dorr R, Zindler G, Preuss A, Stumpfl A. Waiting times and death on the waiting list for coronary artery bypass operation. Experiences in Munich with over 1,000 patients. Herz 1996;21:389 – 96. [3] Ray AA, Buth KJ, Sullivan JA, Johnstone DE, Hirsch GM. Waiting for cardiac surgery: results of a risk-stratified queuing process. Circulation 2001;104(Suppl 1):I92 – 8. [4] Bernstein SJ, Rigter H, Brorsson B, et al. Waiting for coronary revascularization: a comparison between New York State, The Netherlands and Sweden. Health Policy 1997;42:15 – 27. [5] Pitt M, Dutka D, Pagano D, Camici P, Bonser R. The natural history of myocardium awaiting revascularisation in patients with impaired left ventricular function. Eur Heart J 2004;25:500 – 7. [6] Koivula M, Paunonen-Ilmonen M, Tarkka MT, Tarkka M, Laippala P. Fear and anxiety in patients awaiting coronary artery bypass grafting. Heart Lung 2001;30:302 – 11. [7] Naylor CD, Baigrie RS, Goldman BS, Basinski A. Assessment of priority for coronary revascularisation procedures. Revascularisation Panel and Consensus Methods Group. Lancet 1990;335:1070 – 3.

174

F.H.Y. Cesena et al. / International Journal of Cardiology 110 (2006) 167 – 174

[8] Morgan CD, Sykora K, Naylor CD. Analysis of deaths while waiting for cardiac surgery among 29,293 consecutive patients in Ontario, Canada. The Steering Committee of the Cardiac Care Network of Ontario. Heart 1998;79:345 – 9. [9] Jackson NW, Doogue MP, Elliott JM. Priority points and cardiac events while waiting for coronary bypass surgery. Heart 1999; 81:367 – 73. [10] Seddon ME, French JK, Amos DJ, Ramanathan K, McLaughlin SC, White HD. Waiting times and prioritisation for coronary artery bypass surgery in New Zealand. Heart 1999;81:586 – 92. [11] Billing JS, Arifi AA, Sharples LD, Tsui SS, Nashef SA. Heart surgery in UK patients: planned care or crisis management? Lancet 1996;347(9000):540 – 1. [12] Cesena FHY, Favarato D, Cesar LA, de Oliveira SA, da Luz PL. Cardiac complications during waiting for elective coronary artery bypass graft surgery: incidence, temporal distribution and predictive factors. Eur J Cardiothorac Surg 2004;25:196 – 202. [13] Ringqvist I, Fisher LD, Mock M, et al. Prognostic value of angiographic indices of coronary artery disease from the Coronary Artery Surgery Study (CASS). J Clin Invest 1983;71:1854 – 66. [14] Grundy SM, Pasternak R, Greenland P, Smith Jr S, Fuster V. AHA/ACC scientific statement: assessment of cardiovascular risk by

[15]

[16]

[17]

[18]

[19]

use of multiple-risk-factor assessment equations: a statement for healthcare professionals from the American Heart Association and the American College of Cardiology. J Am Coll Cardiol 1999;34: 1348 – 59. Prevention of coronary heart disease in clinical practice. Recommendations of the Second Joint Task Force of European and other societies on coronary prevention. Eur Heart J 1998;19:1434 – 503. Califf RM, Armstrong PW, Carver JR, D’Agostino RB, Strauss WE. 27th Bethesda Conference: matching the intensity of risk factor management with the hazard for coronary disease events. Task Force 5. Stratification of patients into high, medium and low risk subgroups for purposes of risk factor management. J Am Coll Cardiol 1996;27:1007 – 19. Rexius H, Brandrup-Wognsen G, Oden A, Jeppsson A. Mortality on the waiting list for coronary artery bypass grafting: incidence and risk factors. Ann Thorac Surg 2004;77:769 – 74. Knudtson ML. CCS Consensus Conferences. Indications for and access to revascularization. Chapter 9: waiting list issues http://www.ccs.ca, Access: Jan 29, 2005. Rexius H, Brandrup-Wognsen G, Oden A, Jeppsson A. Gender and mortality risk on the waiting list for coronary artery bypass grafting. Eur J Cardiothorac Surg 2004;26:521 – 7.