Progressively increasing operative risk among patients referred for coronary artery bypass surgery.

OBJECTIVE: Advances in surgical, anaesthetic and percutaneous interventional techniques may have led to higher risk patients being referred for coronary artery bypass graft surgery (CABG). The purpose of this study was to compare the predicted mortality risk (EuroSCORE) of a contemporary cohort of patients referred for isolated elective CABG (2002) with that of a cohort referred five years previously (1997) and to examine temporal trends in patient demographics.
METHODS: Records (n=2873) of weekly cardiac surgical referral meetings were examined and the age, sex, type of operation and surgical decision for every patient referred from 1997 to 2002 inclusive were recorded. Furthermore samples of patients referred in 1997 (n=111) and in 2002 (n=110) were chosen, and a complete EuroSCORE was calculated for each patient and compared between groups.
RESULTS: In both 1997 and 2002 the median EuroSCORE among patients not accepted for surgery was significantly higher than those accepted (1997; 3 vs 2, p < 0.001. 2002; 5 vs. 2, p < 0.001). The median EuroSCORE of patients referred in 2002 was significantly higher than those referred in 1997 (3 vs. 2; p < 0.001). There was a progressive increase in median patient age throughout the study period and this accounted for the observed temporal increase in EuroSCORE.
CONCLUSIONS: Predicted mortality risk among patients referred for coronary artery bypass surgery is increasing, mainly due to patient age at referral.

INTRODUCTION

Recent advances in surgical and anaesthetic techniques have been associated with reduced mortality rates from cardiac surgery.1 Improved surgical outcomes, along with major advances in percutaneous coronary intervention2 and changes in the demographics of patients undergoing cardiac catheterisation3 are likely to have resulted in higher risk patients being referred for cardiac surgery compared with previously.4

There are several risk-predicting methods for patients undergoing cardiac surgery.5,6 The EuroSCORE index7,8 gives a practical and numerical prediction of early mortality risk following coronary artery bypass graft surgery (CABG). Its reliability has been demonstrated in clinical practice 9,10 and it is used commonly in the pre-operative assessment process.

The aims of this study were to audit the clinical and demographic profile of patients referred for isolated CABG at our institution, to detect any changes in calculated operative risk (EuroSCORE) over time and to identify factors contributing to any change observed.

METHODS

All patients referred for isolated CABG at Belfast City Hospital were identified from the records of a weekly combined cardiology-cardiac surgery meeting at which all referrals for non-emergency cardiac surgery are made. In all, 2873 patients referred between the beginning of 1997 and the end of 2002 were identified and their age, sex, type of surgery and decision regarding acceptance were collected. Patients referred for valvular surgery were excluded, as were those for whom no decision regarding acceptance was made at the time of discussion.

To study changes in risk profile in detail, a sample of patients was chosen from each of 1997 (n = 111) and 2002 (n = 110). These were the first patients referred in a calendar year for which complete medical records were available and the EuroSCORE was calculated based on identification of relevant patient and cardiac-related factors.5,6 There were no operation-related factors as all patients were scheduled to undergo elective isolated CABG.

Statistical analysis

Variables are presented as median and interquartile range. Between-group analysis used the Mann-Whitney rank sum test. A P value of <0.05 was considered statistically significant.

RESULTS

Activity figures showed a progressive increase in the number of diagnostic coronary angiograms performed annually and a gradual decrease in the percentage referred for elective CABG (Fig 1).

Fig 1

Yearly trends in the number of diagnostic angiograms performed at Belfast City Hospital (blue bars) and the number of patients referred for isolated coronary artery bypass graft surgery (red bars). The solid red line, referring to the right-hand Y axis, represents the percentage of patients undergoing angiography who were referred for isolated CABG

Changes in patient demographics over time

The median age of patients referred for CABG rose progressively between 1997 and 2002 (Table I). Although not reaching statistical significance, the following trends were observed: patients turned down for CABG tended to be older than those accepted in each individual year and patients referred between 2000 and 2002 were more likely to be turned down for CABG.

Table I

Demographic profile of patients referred for coronary artery bypass surgery between 1997 and 2002

Year	All patients referred	Accepted	Rejected	No decision made
1997
Number (%)	420	298 (70.9)	39 (9.3)	83 (19.8)
Median age (IQR)	62 (55-67)	61 (55-67)	63 (57-69)	62 (55-67)
Number (%)female	74 (17.6)	49 (16.4)	11 (28.2)	14 (16.9)
1998
Number (%)	507	368 (72.6)	46 (9.1)	93 (18.3)
Median age (IQR)	62 (55-68)	62 (55-68)	63 (57-70)	61 (55-67)
Number (%)female	97 (19.1)	62 (16.8)	8 (17.4)	27 (29.0)
1999
Number (%)	541	374 (69.1)	59 (10.9)	108 (20.0)
Median age (IQR)	62 (55-69)	61 (55-68)	63 (55-69)	63.5 (54-69)
Number (%)female	112 (20.7)	76 (20.3)	10 (16.9)	26 (24.1)
2000
Number (%)	468	313 (66.9)	106 (22.6)	49 (10.5)
Median age (IQR)	63 (56-69)*	62 (55-68)	65 (59-72)	63 (54-69)
Number (%)female	93 (19.9)	50 (15.9)	31 (29.2)	12 (24.5)
2001
Number (%)	500	323 (64.6)	101 (20.2)	76 (15.2)
Median age (IQR)	63 (57-70)†	63 (57-69)†	65 (57-71)	63 (57-69)
Number (%)female	108 (21.6)	55 (17.0)	31 (30.7)	22 (28.9)
2002
Number (%)	437	286 (65.5)	82 (18.7)	69 (17.8)
Median age (IQR)	65 (58-71)†	64 (58-71)†	66 (59-73)*	68 (59-73) †
Number (%)female	89 (20.4)	55 (19.2)	21 (25.6)	13 (18.8)

p<0.05 versus 1997

p<0.001 versus 1997. IQR = interquartile range

Changes in EuroSCORE over time

The calculated EuroSCORE of the patients sampled from 2002 was significantly higher than those sampled from 1997 (Table II). In both 1997 and 2002 the EuroSCORE among patients turned down for surgery was significantly higher than in patients accepted. In 2002 the median EuroSCORE of all patients referred was similar to that of patients being turned down for surgery in 1997.

Table II

Euroscore values in 1997 and 2002 cohorts. P value column compares accepted with rejected groups

	All referred Median	Accepted	Rejected	p
1997
Number (%)	111	80 (72.1)	31 (27.9)
EuroSCORE	2 (1-3)	2 (0-3)	3 (2.25-4)	<0.001
2002
Number (%)	110	81 (73.6)	29 (26.4)
EuroSCORE	3 (2-5)*	2 (1-4)†	5 (4-8)*	<0.001

Median (interquartile range).

p<0.001 vs. 1997,

p<0.05 vs. 1997.

Further analysis of the observed rise in EuroSCORE between 1997 and 2002 revealed that patient age was the dominant component with almost 100 extra EuroSCORE points being awarded for age in 2002 (Table III). In both 1997 and 2002, the largest age group of patients referred for elective CABG was in the 60-69 year bracket (Fig 2). The second largest group in 1997 was 50-59 years while it was 70-79 years in 2002. In 1997 no patient in the age group 80-89 was referred compared with 4% in 2002. The percentage of females referred for CABG remained similar over time (Table I).

Table III

Components of the EuroSCORE calculation (reference 5) and analysis of the total number of EuroSCORE points awarded for each of these components in samples of patients referred for isolated CABG in 1997 (n = 111 patients) and 2002 (n = 110 patients)

	Definition	Score	1997 cohort	2002 cohort
Patient-related factors
Age	Per 5 years or part thereof over 60 years	1	118 (44.5%)	214 (57.2%)
Gender	Female	1	23 (8.7%)	14 (3.7%)
Chronic pulmonary disease	Long term use of bronchodilators or steroids	1	9 (3.4%)	9(2.4%)
Extra cardiac arteriopathy	See belowa	2	14 (3.4%)	9 (2.4%)
Neurological dysfunction	Disease severely affecting ambulation or day-to-day functioning	2
Previous cardiac surgery	Previous surgery requiring opening of the pericardium	3	9 (3.4%)	3 (1%)
Serum creatinine	≥ 200μmol/L pre-operatively	2	2 (1%)	0
Active endocarditis	Under antibiotic treatment for endocarditis at time of surgery	3
Critical pre-operative state	See belowb	3
Cardiac-related factors
Unstable angina	Requiring iv nitrates until arrival in the operating room	2
LV dysfunction	Moderate (EF 30-50%) Poor <30%	1 or 3	58 (21.9%)	56 (15%)
Recent myocardial infarct	<90 days	2	32 (12.1%)	42 (11.2%)
Pulmonary hypertension	Systolic PA pressure > 60mmHg	2
Operation-related factors
Emergency	Carried out on referral before the beginning of the next working day	2
Other than isolated CABG	Major cardiac operation other than or in addition to CABG	2
Surgery on the thoracic aorta	Ascending, arch or descending aorta	3
Post infarct septal rupture		4

Any of: claudication, carotid occlusion or >50% stenosis, previous or planned surgery on the abdominal aorta, limb arteries or carotids

Ventilation before arrival in the operating room, pre-operative inotropic support, intra-aortic balloon counterpulsation (IABP) or preoperative acute renal failure (anuria or oliguria <10ml/hr). Several patient- and operation-related factors did not apply to this study of elective CABG

Fig 2

Age profiles of patients referred for elective coronary artery bypass surgery in 1997 (red bars) and 2002 (blue bars).

DISCUSSION

Results from this audit indicate a progressive, temporal increase in the risk status of patients referred for CABG as assessed by the EuroSCORE. This increased risk is predominantly a reflection of older patient age. The observations may reflect changes in the population undergoing diagnostic cardiac catheterisation, changes in cardiology practice or advances in cardiac surgical and anaesthetic practice.

The EuroSCORE as a risk predictor for CABG mortality

The EuroSCORE was derived from analysis of 19,030 patients undergoing CABG across Europe.7 It is calculated by additive point scoring based on preoperative patient, cardiac and operation-related factors. Subsequently it has been validated as an accurate predictor of mortality in Europe 9 and North America.10 Based on the EuroSCORE, patients are designated as low risk (EuroSCORE 0-2, average predicted mortality 0.8%), medium risk (EuroSCORE 3-5, predicted mortality 3%) and high risk (EuroSCORE >6, predicted mortality 11.2%).7

When applied to the population in the present study, the average patient referred for elective CABG in 2002 was of medium risk status (median EuroSCORE 3, Table II) compared with low risk status in 1997 (median EuroSCORE 2). The median EuroSCORE of patients accepted for CABG did not change between 1997 and 2002 although the distributions of EuroSCORE values point to a higher risk cohort in 2002. Patients turned down for CABG had substantially higher risk scores compared with those accepted and this disparity increased markedly with time. The data suggest that high-risk patients were more likely to undergo diagnostic cardiac catheterisation and to be referred for elective CABG in 2002 compared with 1997.

Reasons underlying the increase in EuroSCORE over time

The progressive increase in EuroSCORE in the present study is largely due to rising patient age. While populations in the industrialised world are aging, this has not occurred rapidly enough to account for the rise in EuroSCORE we observed. Changes in the epidemiology and treatment of cardiovascular disease may have resulted in patients presenting for the first time later in life.11,12 This may be due to advances in primary prevention, secondary prevention or both. Advances in medical therapy13,14 and in percutaneous coronary intervention 15 may delay or prevent patients' referral for surgery. Patients' and clinicians' expectations of standard care may also have changed from a conservative to a more invasive view, resulting in increased angiography rates in elderly patients.16

As a result, increasingly only those patients who have a survival benefit from surgery17 or those for whom PCI is technically not feasible are being referred for elective CABG. Furthermore the wider availability of PCI compared with CABG may have resulted in some patients undergoing percutaneous treatment who might otherwise have been referred for CABG.

While elderly patients (>70 years) have a higher mortality rate after cardiac surgery than their younger counterparts,16 in selected elderly populations, CABG can achieve excellent improvements in quality of life without excessive mortality, even in those over 80 years.16,18–20 Therefore it is recommended that CABG not be denied on the basis of age alone, especially if elective procedures prevent the subsequent need for emergency surgery with higher associated mortality.

Among the patients we sampled in 1997 and 2002, no risk-predicting factor other than patient age changed over time (Table III). Specifically there was no increase in the numbers of patients being referred with severe LV dysfunction, renal impairment, chronic lung disease or other co-morbidity. The present study focused on patients referred for isolated CABG (approximately 80% of all referrals), and therefore excluded those being referred for valve surgery and emergency surgery, among whom different patterns of co-morbidity may have been observed.

Implications for health care funding

The increase in EuroSCORE we observed has implications for those involved in health care funding. Although the EuroSCORE is designed to predict mortality rather than morbidity, our observations anticipate an increase in the number of patients who might expect a complicated postoperative course and a prolonged intensive care stay.

CONCLUSION

The risk status of patients referred for CABG in 2002 was substantially higher than in 1997, due entirely to differences in patient age. Reasons for this trend are likely to be multifactorial, including changes in preventive medicine and invasive cardiology practice. It is likely that such trends will continue, with implications for cardiologists, cardiac surgeons, intensive care providers and health care purchasers.