HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by van Domburg, R.T. Articles by Bogers, A.J.J.C. Search for Related Content PubMed PubMed Citation Articles by van Domburg, R.T. Articles by Bogers, A.J.J.C. Related Collections Coronary disease

Eur J Cardiothorac Surg 2002;21:733-740
© 2002 Elsevier Science NL

Short-term and 5-year outcome after primary isolated coronary artery bypass graft surgery: results of risk stratification in a bilocation center

Department of Cardiothoracic Surgery, University Hospital Rotterdam Dijkzigt, Thoraxcenter, Location 5 Midden Room H539, Dr. Molewaterplein 40, Erasmus MC, 3015 GD Rotterdam, The Netherlands

Received 19 October 2001; received in revised form 9 January 2002; accepted 15 January 2002.

* Corresponding author. Tel.: +31-10-463-3933; fax: +31-10-408-9484
e-mail: vandomburg{at}thch.azr.nl

	Abstract

Top Abstract 1. Introduction 2. Materials and methods 3. Results 4. Discussion 5. Conclusion Appendix A. Parsonnet score Appendix B. EuroSCORE References

 
Objective: We retrospectively investigated the short and mid-term outcome of non-emergent primary isolated coronary artery bypass graft (CABG) surgery in relation to risk stratification in the fully equipped university location (FE) and the low volume, limited facility location (LVLF) of our department. Methods: Between September 1995 and December 1996, 832 patients were referred to our department to undergo a primary isolated CABG operation. The surgical team selected 482 patients (58%) as being at low-risk. These were treated in the LVLF hospital. The other 350 patients with mixed-risk were treated in the FE hospital. The selection consisted primarily of exclusion of patients with moderate or poor left ventricular function, severe COPD or renal impairment, from surgery in the LVLF location. Finally, the prognostic value of the EuroSCORE and the Parsonnet score was tested on our patient population. Results: Overall in-hospital mortality was 1.6% (13 patients). One patient died in the LVLF group (0.2%) and 12 patients (3.4%) in the FE group. LVLF patients experienced less complications during the hospital period compared to the FE patients (5 versus 21%; P=0.0001). The Parsonnet risk model and the EuroSCORE risk model showed both a good relation with in-hospital mortality. After discharge, an increased risk of late mortality was observed up to 1 year postoperative in the FE group compared to the LVLF group (2.7 versus 0.5%; P=0.01). Risk factors for 5-year mortality were pre-operative renal impairment (blood creatinine >150 µmol/l) (hazard ratio (HR): 2.8; 95% confidence interval (CI): 1.4–5.5), diabetes (HR: 2.1; 95% CI: 1.3–3.5), impaired LVEF (HR: 1.9; 95% CI: 1.2–3.0), COPD (HR: 1.9; 95% CI: 1.1–3.5) and older age (HR: 1.07 per year; 95% CI: 1.01–1.10). Lipid-lowering therapy was a predictor of lower mortality at 5-years (HR: 0.5; 95% CI: 0.4–0.9). Conclusion: By careful decision making, selection of low-risk patients for a low volume and limited facility location resulted in excellent in-hospital survival with very low complication rates.

Key Words: Coronary artery bypass graft • Hospital mortality • Risk assessment • Prognosis

	1. Introduction

Top Abstract 1. Introduction 2. Materials and methods 3. Results 4. Discussion 5. Conclusion Appendix A. Parsonnet score Appendix B. EuroSCORE References

 
Coronary artery bypass surgery (CABG) was first carried out more than 30 years ago [1] and rapidly became a routine operation for revascularization in patients with coronary artery disease. To meet the demands of such sudden increase, our cardiothoracic surgery team performs CABG in two locations. The primary location is the university hospital, which is fully equipped (FE) for all cardiothoracic surgery. In the second limited facility location a low volume, limited facility location program (LVLF) of weekly 6–8 non-emergent primary isolated coronary bypass operations are performed by members of the surgical staff of the university hospital, but only in patients who are selected as being at low-risk. No other cardiac surgery is done at this site. In order to investigate the results of CABG in our practice we followed 832 consecutive patients who underwent CABG surgery at these two locations and followed these patients up to 5 years. We determined the 5-year outcome with regard to mortality and coronary reinterventions and investigated potential predictors of both early and mid-term mortality.

The use of risk assessment models is now widespread. The Parsonnet risk stratification system [2] was developed in the USA in the 1980s and recently the European System for Cardiac Operative Risk Evaluation (EuroSCORE) has emerged [3]. The applicability of the Parsonnet score and the EuroSCORE models was tested retrospectively to our study population.

	2. Materials and methods

Top Abstract 1. Introduction 2. Materials and methods 3. Results 4. Discussion 5. Conclusion Appendix A. Parsonnet score Appendix B. EuroSCORE References

 
Between September 1, 1995 and December 31, 1996 832 consecutive patients were referred to undergo a first non-emergent (elective or semi elective) primary isolated CABG operation (CABG without other open heart procedures during the same hospital admission). The 52 redo CABG operations that were performed in the FE location in this time frame were excluded. Also excluded were 23 patients referred for a non-elective procedure such as failed angioplasty or those who suffered an acute myocardial infarction in the 24 h prior to CABG, all of these patients were treated in the FE location. Of the remaining 832 patients, 482 (58%) were selected as being at low-risk and treated in the LVLF location. Low-risk patients were selected primarily by absence of the following risk factors: moderate or poor left ventricular function, severe COPD or elevated creatinine levels. However, no further strict decision rules or existing risk scores were used in this cohort. Baseline characteristics included the clinical profile, angina and functional status, risk factors and information on coronary angiographic features.

2.1. Fully equipped location versus low volume limited facility location
The cardiothoracic surgeon and perfusionist from the FE location performed the surgery, assisted by residents or a fellow in general surgery from the LVLF location. Limited material facilities and no prosthetic or valvular material are at hand in the LVLF location. After surgery the cardiothoracic surgeon is not present in the LVLF unit, but on call. There are no specific residents available. Patients stay in a general intensive care unit, seen by anesthesiologists. The ward is incorporated in the department of cardiology. The limited outpatient clinic is incorporated in the department of general surgery. If problems arise that are beyond the LVLF unit's abilities, sometimes the necessary expertise and material is transported from the FE location to the LVLF location, but mostly the patient is transported to the FE location for further treatment.

2.2. Surgical technique
Standard extracorporeal circulation with moderate hypothermia and cardioplegic arrest (St. Thomas solution) was used for the operations. An average of 2.8 distal coronary anastomoses per patient was made.

2.3. Follow-up
At the time of the follow-up, June 1 2001, follow-up status was obtained retrospectively by review of the hospital records and checked against the civil registries. Furthermore, general practitioners were approached for recording of myocardial infarctions, strokes and possible subsequent revascularizations that were not performed in our center. If no further information from the general practitioner could be obtained, the patients were approached after retrieval of vital status. The cause of death was determined by checking our hospital records or by contacting the referring hospitals or the treating general practitioner. With regard to the cause of death a distinction was made between cardiac death (sudden death <1 h of complaints, fatal acute myocardial infarction, congestive heart failure) and non-cardiac death (all other causes). In eight of the 92 patients who died (9%) the cause of death could not be retrieved. In three patients who had moved abroad, survival status could not be retrieved and the last available follow-up data were used. Follow-up was complete in all other patients (99.5%).

2.4. Risk assessment
We tested the established Parsonnet [2] and the EuroSCORE [3] risk assessment models on our study population (Appendix). The Parsonnet model allocates additive predicted mortality percentage points to 14 patient risk factors to give a Parsonnet score, which is supposed to be indicative of the percent mortality for each patient. For the EuroSCORE model 68 preoperative risk factors and 29 operative variables were collected in 128 participating centers in 1995. This resulted in nine patient-related factors, four cardiac-related factors and four operation-related factors.

2.5. Statistical methods
Continuous data were expressed as mean value±standard deviation (SD) and were compared using Student's t-test, categorical variables by chi-square tests. Cumulative survival curves were constructed using the Kaplan–Meier method. Among patient subgroups the log-rank test was used to compare survival curves. A logistic multivariable regression, stratified by the two hospitals, was used to adjust for the baseline differences to determine the impact on in-hospital mortality and hospital complications. Odds ratios and 95% confidence intervals were calculated to test for significant differences in early events between the two hospitals.

The following events were classified as in-hospital complications: acute myocardial infarction, cerebro vascular accident, prolonged artificial ventilation (>2 days), reanimation, rethoracotomy for bleeding, intra aortic balloon pump or reoperation for wound complication. To determine independent predictors of 1- and 5-year survival multivariate analysis was carried out using stepwise (forward and backward) Cox regression. Pre-selected variables were age, gender, diabetes (defined as pharmacologically treated; types I or II), treated hypertension, the use of medication for hypercholesterolemia, number of diseased vessels (stenosis >50%), qualitative left ventricular ejection fraction (normal if >55%; reduced if >45 and 55%; moderate if >30% and 45%; poor if 30%), chronic obstructive pulmonary disease (COPD; pharmacologically treated) and renal impairment (blood creatinine >150 µmol/l).

	3. Results

Top Abstract 1. Introduction 2. Materials and methods 3. Results 4. Discussion 5. Conclusion Appendix A. Parsonnet score Appendix B. EuroSCORE References

 
In Table 1 preoperative patient characteristics are displayed for all patients, and separately for the FE location and the LVLF location. LVLF patients were almost 2 years younger compared to FE patients (63.8 years versus 62.0 years; P=0.01). The LVLF group contained relatively more patients with two vessel disease (28 versus 19%) and less patients with three vessel disease (56 versus 68%; P=0.01) and most patients had a normal ejection fraction (91 versus 47%; P=0.0001). Furthermore, fewer patients in the LVLF group had diabetes (13 versus 18%; P=0.04) and prior myocardial infarctions (44 versus 57%; P=0.001). Table 2 displays the procedural data of both hospitals. The surgical techniques were similar in both centers. Arterial grafts were used in 90% of the patients. The left internal mammary artery (LIMA) was used to graft the left anterior descending artery (LAD) in 75% of the patients. The right internal mammary artery (RIMA) was used in 6% and the gastric epiploica artery (GEA) in 3% of the patients. A combination of vein grafts and arterial grafts were used in the majority of the patients (81%). Exclusively vein grafts were used in 10% of the patients and exclusively arterial grafts in 9% of the patients. The main reasons for not using arterial grafts were the physical condition of the patients (26%), severe unstable angina (15%), poor left ventricular function (11%), renal failure (11%) or a left ventricular aneurysm (9%). The right coronary artery (RCA) and left circumflex (LCX) were grafted more often in the FE group because of the more extensive vessel disease in this patient group as evidenced by more distal anastomoses in this group (2.9 versus 2.7%; P=0.01). The T-graft procedure was used in 7% of the patients. Endarterectomy was performed in 15 FE patients of whom three died. One patient had a complicated and extensive postoperative period (2 months post-operative artificial ventilation) and died after 3 months. The other two patients died after 4 years from non-cardiac causes. All four LVLF patients in whom endarterectomy was performed were still alive at the end of the follow-up.

View larger version (19K): [in this window] [in a new window]   Fig. 1. Distribution of the Parsonnet score according to the fully equipped location (FE) and low volume, limited facility location (LVLF) (left Y-axis). The dotted line represents the relation of the Parsonnet score with the in-hospital mortality (right Y-axis).

View larger version (20K): [in this window] [in a new window]   Fig. 2. On the left Y-axis the distribution of the EuroSCORE according to the FE location and LVLF location is shown. The dotted line represents the relation of the EuroSCORE score with the in-hospital mortality (right Y-axis).

View larger version (14K): [in this window] [in a new window]   Fig. 3. Cumulative Kaplan–Meier survival curves according to the FE location and LVLF location. The vertical lines represent the 95% confidence intervals (±2 SD). The dotted line represent the survival curve of the normal Dutch population matched for age and sex.

View larger version (12K): [in this window] [in a new window]   Fig. 4. Cumulative Kaplan–Meier event-free survival curves (freedom from death/MI/repeat CABGPTCA or CVA) according to the fully equipped location (FE) and LVLF location. The vertical lines represents the 95% confidence intervals (±2 SD).

Independent risk factors of increased 5-year mortality were renal impairment (HR: 2.8; 95% CI: 1.4–5.5), diabetes (HR: 2.1; 95% CI: 1.3–3.5), impaired left ventricular ejection fraction (HR: 1.9; 95% CI: 1.2–3.0), COPD (HR: 1.9; 95% CI: 1.1–3.5) and higher age (HR: 1.07 per year; 95% CI: 1.01–1.10). Already at 5 years post-CABG it is evident that the use of only vein grafts is probably related to higher mortality (HR: 1.9; 95% CI: 0.95–3.8). Therapy for hyperlipidemia was an independent predictor of lower mortality (HR: 0.6; 95% CI: 0.4–0.9) (Table 4).

	4. Discussion

Top Abstract 1. Introduction 2. Materials and methods 3. Results 4. Discussion 5. Conclusion Appendix A. Parsonnet score Appendix B. EuroSCORE References

Despite large differences in patient-profile, the in-hospital mortality rate was similar compared to earlier series from our department of patients who underwent CABG in the 1970s (1.6 now versus 1.2% in the 1970s) [4]. Compared with other registries our in-hospital mortality results are below average. Lewsey et al. reported the Scottish Registry [5] results and found 2.7% mortality at 30-days in a similar patient subpopulation and Hannan et al. from the large New York Registry [6] reported an in-hospital mortality of 1.9% after primary elective CABG.

Only one of the 482 LVLF patients died before discharge and in only 5% complications occurred during admission. Only experienced surgeons were operating in the LVLF location. Although our group at that time did not use any established risk model, the decision to allocate the patients to the low-risk group, which could be referred to the LVLF location, turned out to be clinically very applicable.

Due to hospital strategies at that time (1995) and albeit the higher risk profile of the FE location patients, hospital stay was in both locations similar (11 days). Since then, the median hospital stay has decreased gradually.

4.1. Risk assessment models for in-hospital mortality
Jones et al. [7] defined and prioritized clinical variables to reflect their importance for relating to in-hospital mortality after CABG and concluded that only seven clinical variables already provided a large amount of prognostic information. Bridgewater et al. showed [8] that risk scores from the USA, or even from other European countries might have only limited predictive ability. Sergeant et al. [9] used a learning set from his own center of patients who operated between 1971 and 1987, and tested his formula on patients who operated between 1987 and 1992. A limitation of this model is that over time clinical criteria are changing and that the operations and knowledge are evolving. Retrospectively, we tested the Parsonnet score [2] and EuroSCORE [3] on our patient population. Both scores showed a good relation with in-hospital mortality. However, the total number of events was too low to draw conclusions. Therefore, we used both models only to illustrate the patients’ risk profile.

4.2. Long-term outcome
We observed that patients from the FE location had an increased risk of mortality after hospital discharge the first postoperative year (n=10; 2.7%) (Fig. 3). This finding is in agreement with Noyez et al. [10] who concluded that for risk assessment the in-hospital period is not sufficient, but should be expanded to at least 6 months. In contrast, the risk of mortality in the LVLF location patients remained constant directly after discharge throughout the end of the follow-up period (0.5%/year).

The mid-term prognosis up to 5 years was excellent. The yearly mortality rate after 1 year follow-up was only 1.5% in patients treated at the LVLF location and 2% for the FE patients, and this is similar to the normal Dutch population matched for age and sex (1.7%) (Fig. 3). Coronary reintervention was rare in the first 5 years. According to an earlier report from our institution [4] higher intervention rates can be expected after the 8 years, mostly due to vein graft restenosis.

4.3. Predictors of mortality
In concordance with previous studies, we found that renal impairment, diabetes (types I or II), impaired left ventricular function, COPD and age were independent risk factors for higher mortality. Additionally to these findings hyperlipidemia was also an independent predictor of lower mortality. Patients with hyperlipidemia are normally at risk for higher mortality. Surprisingly, we found that in patients who were treated for hyperlipidemia this risk was not only diminished but also even changed into a significantly lower risk for mortality. This illustrates a significant under recognition and under treatment of hyperlipidemia.

	5. Conclusion

Top Abstract 1. Introduction 2. Materials and methods 3. Results 4. Discussion 5. Conclusion Appendix A. Parsonnet score Appendix B. EuroSCORE References

 
Although no formal risk model for primary CABG was used, careful stratification for non-emergent treatment in a low volume, limited facility location is feasible. Using this strategy in-hospital mortality and the complication rate were low.

	Acknowledgments

 
The support of Mrs Ada Matser-van der Berg and Mrs Marijke Rozema-Botermans on the data collection is highly appreciated. This study would not have been possible without the kind support of many general practitioners.

	Appendix A. Parsonnet score

Top Abstract 1. Introduction 2. Materials and methods 3. Results 4. Discussion 5. Conclusion Appendix A. Parsonnet score Appendix B. EuroSCORE References

 

Risk factor

Score

Female

1

Obesity (>1.5 ideal weight)

3

Diabetes

3

Hypertension

3

Ejection fraction

Normal

0

Reduced or moderate

2

Poor

4

Age (years)

70–74

7

75–80

12

80

20

Dialysis dependent

10

By selection, all other risk factors (reoperation, preoperative IABP, LV aneurysm, emergency from cathlab, catastrophic states and valve surgery) used in the Parsonnet score, were absent in our study population.

	Appendix B. EuroSCORE

Top Abstract 1. Introduction 2. Materials and methods 3. Results 4. Discussion 5. Conclusion Appendix A. Parsonnet score Appendix B. EuroSCORE References

 

Age (per 5 over 60 years)

1

Female

1

COPD

1

Renal impairment (creatine>150 µmol/l)

2

Unstable angina

2

Reduced or moderate ejection fraction

1

Poor ejection fraction

3

Recent myocardial infarction (<90 days)

2

Postinfarct septal rupture

4

By selection, all other risk factors (extracardiac arteriopathy, neurological dysfunction disease, previous cardiac surgery, endocarditis, critical preoperative state, emergency, other than isolated CABG and surgery on thoracic aorta) used in the EuroSCORE score, were absent in our study population. Pulmonary hypertension was not known in our population and thus ignored.

COPD=chronic obstructive pulmonary disease

	References

Top Abstract 1. Introduction 2. Materials and methods 3. Results 4. Discussion 5. Conclusion Appendix A. Parsonnet score Appendix B. EuroSCORE References

 

1. Favaloro R.G. Saphenous vein graft in the surgical treatment of coronary artery disease. Operative technique. J Thorac Cardiovasc Surg 1969;58:178-185.[Medline]
2. Parsonnet V., Dean D., Bernstein A.D. A method of uniform stratification of risk for evaluating the results of surgery in acquired adult heart disease. Circulation 1989;79:I3-12.
3. Nashef S.A., Roques F., Michel P., Gauducheau E., Lemeshow S., Salamon R. European system for cardiac operative risk evaluation (EuroSCORE). Eur J Cardiothorac Surg 1999;16:9-13.[Abstract/Free Full Text]
4. Veldkamp R.F., Valk S.D., van Domburg R.T., van Herwerden L.A., Meeter K. Mortality and repeat interventions up until 20 years after aorto-coronary bypass surgery with saphenous vein grafts. A follow-up study of 1041 patients. Eur Heart J 2000;21:747-753.[Abstract/Free Full Text]
5. Lewsey J.D., Murray G.D., Leyland A.H., Boddy F.A. Comparing outcomes of percutaneous transluminal coronary angioplasty with coronary artery bypass grafting; can routine health service data complement and enhance randomized controlled trials?. Eur Heart J 1999;20:1731-1735.[Abstract/Free Full Text]
6. Hannan E.L., Racz M.J., McCallister B.D., Ryan T.J., Arani D.T., Isom O.W., Jones R.H. A comparison of 3-year survival after coronary artery bypass graft surgery and percutaneous transluminal coronary angioplasty. J Am Coll Cardiol 1999;33:63-72.[Abstract/Free Full Text]
7. Jones R.H., Hannan E.L., Hammermeister K.E., Delong E.R., O'Connor G.T., Luepker R.V., Parsonnet V., Pryor D.B. Identification of preoperative variables needed for risk adjustment of short-term mortality after coronary artery bypass graft surgery. The Working Group Panel on the Cooperative CABG Database Project. J Am Coll Cardiol 1996;28:1478-1487.[Abstract]
8. Bridgewater B., Neve H., Moat N., Hooper T., Jones M. Predicting operative risk for coronary artery surgery in the United Kingdom: a comparison of various risk prediction algorithms. Heart 1998;79:350-355.[Abstract/Free Full Text]
9. Sergeant P., Blackstone E., Meyns B. Can the outcome of coronary bypass grafting be predicted reliably?. Eur J Cardiothorac Surg 1997;11:2-9.[Abstract]
10. Noyez L., Verheugt F.W., Peppelenbosch A.G., Skotnicki S.H., Brouwer M.H. Aortocoronary bypass surgery; at least 6 months follow-up required for assessment of postoperative course. Ned Tijdschr Geneeskd 2000;144:1874-1877.[Medline]

This article has been cited by other articles:

				U. M. Idanpaan-Heikkila, L. Lambie, S. Mattke, V. McLaughlin, H. Palmer, and J. V. Tu Selecting indicators for the quality of cardiac care at the health system level in Organization for Economic Co-operation and Development countries Int. J. Qual. Health Care, September 1, 2006; 18(suppl_1): 39 - 44. [Abstract] [Full Text] [PDF]

				B. J. Leavitt, C. S. Ross, B. Spence, S. D. Surgenor, E. M. Olmstead, R. A. Clough, D. C. Charlesworth, R. S. Kramer, G. T. O'Connor, and for the Northern New England Cardiovascular Diseas Long-Term Survival of Patients With Chronic Obstructive Pulmonary Disease Undergoing Coronary Artery Bypass Surgery Circulation, July 4, 2006; 114(1_suppl): I-430 - I-434. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by van Domburg, R.T. Articles by Bogers, A.J.J.C. Search for Related Content PubMed PubMed Citation Articles by van Domburg, R.T. Articles by Bogers, A.J.J.C. Related Collections Coronary disease