Background. The Mayo Clinic Risk Score (MCRS) is a validated numeric score that predicts outcome following primary percutaneous coronary intervention (PCI). Purpose. We evaluated the ability of MCRS to risk stratify patients undergoing primary angioplasty. Methods. Patients undergoing primary angioplasty within 6 hours of the onset of chest pain in the New York State percutaneous coronary intervention reporting system (n = 3,005) had their MCRS calculated using predictive variables: age, presence of cardiogenic shock, renal failure, class III/IV congestive heart failure, left main coronary disease and multivessel coronary disease. All patients were presumed to have intra-coronary thrombus and undergoing an urgent/emergent procedure. Based on the MCRS, patients were classified into five risk categories: very low risk (MCRS < 5), low risk (6ñ8), moderate (9ñ11), high (12ñ14) and very high risk (15ñ25). Results. The mean age of the study population was 62 years, 70% were male; stents were used in 89% and glycoprotein IIb/IIIa antagonists in 72%. The prevalence of cardiogenic shock, multivessel disease and left main disease was higher in patients with MCRS > 12. Overall in-hospital mortality following primary angioplasty was 4.7%; it was 0% in the ìvery low riskî category, 0.9% in the ìlow riskî category, 3.2% in the ìmoderate riskî category, 10.7% in the ìhigh riskî category, and 25.1% in the ìvery high riskî category (p < 0.0001). The higher-risk MCRS category predicted increased risk even when 317 (10.5%) patients with cardiogenic shock were excluded from the analysis. The overall c-statistic for the prediction of in-hospital mortality by MCRS was 0.85. Conclusion. Increasing MCRS predicts in-hospital mortality following primary angioplasty.

       Current ACC/AHA guidelines support the use of percutaneous coronary intervention (PCI) as reperfusion therapy for ST elevation myocardial infarction (STEMI) when performed in a timely manner by experienced operators.¹ Retrospective analysis suggests that primary angioplasty may be preferable to thrombolysis as a mode of reperfusion as the estimated risk of mortality increases.² While existing models such as the TIMI Risk Score³ and PAMI Risk Score⁴ can be used to estimate risk prior to thrombolysis and primary angioplasty respectively, their utility may be limited as high-risk patients such as those in cardiogenic shock, renal failure, those ineligible to receive fibrinolysis were excluded in these studies and as the models did not include procedural/angiographic characteristics.

Figure 1

The Mayo Clinic Risk Score is calculated by adding the integer scores corresponding to the risk factors in the variable column. Based upon the Integer Score, the individual patientís risk category can be determined. Adapted from <i>J Am Coll Cardiol 2002;40;387ñ393</i>.

       Several models that include both clinical predictors and procedural characteristics have been described and predict hospital outcomes in patients undergoing PCI.^5ñ7 The Mayo Clinic Risk Score (MCRS) is one such model that is calculated using clinical and angiographic predictors7 (Figure 1). Increasing MCRS has been correlated with higher in-hospital adverse events in different datasets.^7,8 Since several variables in the MCRS are independent predictors of mortality following primary angioplasty,⁹ we hypothesized that MCRS could serve as a risk prediction tool for mortality in such patients. Therefore, the purpose of this investigation was to determine whether the MCRS predicted in-hospital mortality among patients undergoing primary angioplasty in the New York State PCI Reporting System (PCIRS) database.

Methods
       Database. The PCIRS database is a mandatory statewide registry that captures specific clinical and procedural characteristics of all patients undergoing PCI and records their in-hospital outcomes. It includes data from all centers performing PCI in New York State and was developed in 1991 by the New York State Department of Health (DOH) and the Cardiac Advisory Committee (CAC). The responsible physician or designee completes a questionnaire that is entered into a computer database at the Department of Health. The information includes patient demographics, risk factors, procedural, angiographic characteristics and in-hospital outcomes. The data are audited periodically for errors, and discrepancies brought to the attention of the institutions for rectification.
       Patient population. Patients undergoing PCI for Acute Myocardial Infarction (AMI) within six hours of symptom onset, excluding those receiving thrombolysis within seven days, were defined as undergoing primary angioplasty. This definition of primary angioplasty has been accepted and used in prior investigations using the PCIRS database.^5,10

Figure 2

In-hospital mortality increases as Mayo Clinic Risk Score (MCRS) increases, especially above a MCRS of 12.

       Mayo Clinic Risk Score derivation. The MCRS was derived using previously published criteria.⁷ It was calculated as a sum of the weighted integer scores based upon the presence of the characteristics shown in Figure 1. Based upon the sum risk score, patients were further classified into risk categories of very low risk (MCRS = 0ñ5), low risk (MCRS = 6ñ8), moderate risk (MCRS = 9ñ11), high risk (MCRS = 12ñ14) and very high risk (MCRS = 15 and above).
       Since primary angioplasty is by definition an emergent procedure and pathophysiologically associated with a thrombus in nearly all patients, we assumed that these two conditions were fulfilled in all patients. In the PCIRS, hemodynamic instability was defined as the requirement for pharmacological and mechanical support to maintain blood pressure or cardiac output prior to angioplasty. Acute hypotension (systolic BP < 80 mmHg) or low cardiac index (< 2.0 l/min.m2) despite pharmacologic and/or mechanical support was defined as cardiogenic shock. For purposes of calculating the MCRS, patients with hemodynamic instability or cardiogenic shock were considered to be in shock (n = 317; 10.5%). We defined renal failure as either serum creatinine > 2.5 mg/dl or renal failure requiring dialysis. Vessel disease was defined as the presence of at least one lesion with > 70% diameter stenosis. Angiographic success was defined as the successful dilation of a lesion with the achievement of a final post-procedural stenosis less than 50%. Procedural success was defined as ìnoneî if none of the attempted lesions achieved angiographic success; ìpartialî if at least one but not all attempted lesions achieved angiographic success; or ìtotalî if all attempted lesions achieved angiographic success.
       Statistical analysis. The available data were imported into and analyzed using SPSS 12.0 (SPSS, Inc.) statistical analysis program. Continuous variables are presented as mean (± standard deviation) and categorical variables as percentages. Continuous variables were compared using Studentís t-test and categorical variables using chi-square test. The analysis of the comparison of outcomes by the MCRS was performed using a Chi-square statistic with Cochran-Armitage test for trend. A logistic regression model11 of mortality was constructed using the MCRS as explanatory variable. The model discrimination was assessed using the area under the receiver operating curve (ROC) or the c-statistic. Model goodness-of-fit was assessed using the Hosmer-Lemeshow test and fit considered adequate if computed p > 0.05. All comparisons are two-sided and statistical significance claimed at a computed p-value < 0.05.

Results
       Of the 74,969 patients in the PCIRS, 3005 (4%) underwent primary angioplasty within six hours of symptom onset. Baseline characteristics of all patients who underwent primary angioplasty are shown in Table 1. Overall, the mean age of the patient population was 62 ± 13 years, 70% were male. Renal failure was observed in 1.9%, congestive heart failure on admission in 10.7%, hemodynamic instability in 6.8% and cardiogenic shock in 3.8%. Single vessel coronary artery disease (CAD) was observed in 56% of patients and left main CAD in 2.1%. The mean ejection fraction calculated by ventriculography or echocardiography was 46% ± 12%. Complete procedural success was achieved in 93% with stents used in 89% and glycoprotein IIb/IIIa inhibitors in 72%.
       MCRS was calculated and patients categorized into five risk categories as shown in Figure 1. The distribution of the variables within each risk category is shown in Table 2. In the very low-risk category, all patients were < 40 years of age with single vessel disease. Patients classified as low risk had predominantly single vessel CAD and few co-morbidities without any cardiogenic shock. The presence of renal failure, congestive heart failure and multivessel disease was higher among patients classified at moderate risk and above. Left main CAD was present more often in the high and very high-risk groups. In general, most patients classified as very high risk demonstrated a combination of cardiogenic shock, multivessel and/or left main CAD.
       When in-hospital mortality was plotted against MCRS (Figure 2), it increased linearly as the score increased. Patients with risk score < 12 (n = 2,461; 89.9%) experienced lower in-hospital mortality compared to those with risk scores > 12 (n = 544; 18.1 %); in-hospital mortality 2% versus 16.7%, p < 0.0001. As seen in Figure 2, patients with very low risk experienced 0% in-hospital mortality, whereas it was 0.9% in those at low risk, 3.2% in moderate risk, 10.7% in high risk and 25.1% in very high risk category (p < 0.0001, for trend). The composite outcome of death, Q-wave MI, stroke and emergency CABG was observed in 1.4%, 1.6 %, 4.9%, 12% and 27.3%, respectively, among the risk categories (p = < 0.0001, Table 3). Since cardiogenic shock is a powerful predictor of mortality during AMI in general and contributed significantly to total mortality within this dataset, we examined MCRS performance after excluding patients in cardiogenic shock from analysis (Figure 3). This resulted in the exclusion of 117 (37%) patients in the high-risk group and 181 (79%) the very high risk groups. Despite the exclusion of these patients, those classified as high risk and very high risk experienced in-hospital mortality of 11% and 19.6% respectively. In logistic regression analysis of in-hospital mortality using MCRS variables as the only explanatory variables, the c-statistic was 0.85, suggesting excellent discriminative ability of MCRS in this population.

Figure 3

In-hospital mortality by Mayo Clinic Risk Score categories for all patients (solid bars) and in those without cardiogenic shock (empty bars). In-hospital mortality increases progressively from very low risk to very high risk categories regardless of the presence or absence of cardiogenic shock.

Discussion
       We applied the Mayo Clinic Risk Score to patients undergoing primary angioplasty in the PCIRS and observed that increasing MCRS was associated with a higher risk for mortality, independent of the presence of cardiogenic shock. In particular, patients with a MCRS > 12 experienced over eight times the in-hospital mortality compared to those with MCRS < 12.
       Whereas several risk prediction models have been described that identify in-hospital outcomes following both STEMI³ and primary angioplasty,^4,5 their clinical use has not been widespread due to inherent limitations. For instance, in the prediction model using the 1995 New York State angioplasty database, contemporary therapies such as intracoronary stents and adjunct glycoprotein IIb/IIIa inhibitor therapy was used infrequently. These advances in percutaneous revascularization have improved clinical outcomes,¹² thereby limiting the applicability of the model derived when balloon angioplasty was the dominant mode of catheter-based therapy.
       The MCRS is a validated risk prediction model⁸ that has demonstrated better prediction of PCI outcomes than the ACC lesion classification.¹³ In the MCRS model, four out of the eight predictor variables (cardiogenic shock, Class 3 or greater CHF, emergency procedure and visible thrombus) are often observed in the setting of AMI.^9,14 Although AMI accounted for a small number of patients in the original datasets, these AMI related variables contributed more weight to MCRS7 likely explaining why MCRS performed extremely well by identifying a gradient of risk following primary angioplasty when applied in the PCIRS.
       Increasing MCRS was associated with higher in-hospital mortality (Figure 2), especially above a risk score of 12. Several reasons may be advanced to account for the predictive ability of MCRS in this patient population. While out of hospital mortality during AMI is a result of malignant ventricular arrhythmias^1,5 in-hospital mortality is largely determined by the extent of myocardial injury and the subsequent left ventricular dysfunction.^16,17 This response to the myocardial insult may be modulated by the age of the patient, the extent of coronary disease and the presence of comorbidities.⁵ While patients with extensive infarcts are more likely to experience hemodynamic instability, cardiogenic shock and Class 3 heart failure, older patients with co-morbidities may experience similar risk even with smaller infarcts.^18,19 It is likely that higher MCRS is an integrated measure of age, existing comorbidities, infarct extent and its hemodynamic consequence during primary angioplasty. Thus younger patients with single vessel CAD, few comorbidities and not in cardiogenic shock had lower MCRS compared to older patients with any combination of multivessel CAD, left main CAD, multiple comorbidities and those in cardiogenic shock.
       Whereas cardiogenic shock is an established, powerful predictor of mortality despite aggressive revascularization therapy,²⁰ we observed that, even in the absence of cardiogenic shock, a MCRS score above 12 predicted higher risk (Figure 3) indicating that the risk assessment by MCRS extends beyond the mere recognition of cardiogenic shock. The c-statistic measure of 0.85 during primary angioplasty compares favorably to the original c-statistic of 0.78 observed among patients undergoing emergent and urgent procedures in the original dataset.⁸
       However, certain limitations must be acknowledged with this analysis. Since the MCRS was originally derived to compare PCI outcomes in all patients, it could be argued that the combination of variables, their coefficients and the subsequent score may have been different if derived from patients with AMI alone. In the original database, patients with renal failure were identified if dialysis was needed or if the serum creatinine was > 3.0 mg/dl, whereas we used a creatinine > 2.5 or the need for dialysis as a marker of renal failure as per data available in the PCIRS. Despite these differences, the definitions used in the current analysis potentially served to classify patients to higher risk category, and may have underestimated the true magnitude of risk. Finally, since the risk score requires both clinical and angiographic information for accurate assessment of risk, its utility during initial decision-making may be limited.

Conclusion
       Despite these limitations, our results suggest that the Mayo clinic risk score could serve as a useful tool to identify risk among patients undergoing primary angioplasty. A prospective evaluation of the Mayo clinic risk score in the setting of primary angioplasty may be warranted.

1. Antman EM, Armstrong PW, Bates ER, et al. ACC/AHA guidelines for the management of patients with ST-elevation myocardial infarction: Executive summary: A report of the ACC/AHA Task Force on Practice Guidelines (Committee to Revise the 1999 Guidelines on the Management of Patients With Acute Myocardial Infarction.) J Am Coll Cardiol 2004;44:671ñ719.
2. Kent DM, Schmid CH, Lau J, Selker HP. Is primary angioplasty for some as good as primary angioplasty for all? J Gen Intern Med 2002;17:887ñ889
3. Morrow DA, Antman EM, Parsons L. et al. Application of the TIMI risk score for ST-elevation MI in the National Registry of Myocardial Infarction 3. JAMA 2001;286:1356ñ1359.
4. Addala S, Grines CL, Dixon SR et al. Predicting mortality in patients with ST-elevation myocardial infarction treated with primary percutaneous coronary intervention (PAMI risk score). Am J Cardiol 2004;93:629ñ632.
5. Hannan EL, Racz MJ, Arani DT et al. Short- and long-term mortality for patients undergoing primary angioplasty for acute myocardial infarction. J Am Coll Cardiol 2000;36:1194ñ1201.
6. Qureshi MA, Safian RD, Grines CL et al. Simplified scoring system for predicting mortality after percutaneous coronary intervention. J Am Coll Cardiol 2003;42:1890ñ1895.
7. Singh M, Lennon RJ, Holmes DR et al. Correlates of procedural complications and a simple integer risk score for percutaneous coronary intervention. J Am Coll Cardiol 2002;40:387ñ393.
8. Singh M, Rihal CS, Selzer F et al. Validation of Mayo Clinic risk adjustment model for in-hospital complications after percutaneous coronary interventions, using the National Heart, Lung, and Blood Institute dynamic registry. J Am Coll Cardiol 2003;42:1722ñ1728.
9. Block PC, Peterson ED, Krone R et al. Identification of variables needed to risk adjust outcomes of coronary interventions: Evidence-based guidelines for efficient data collection. J Am Coll Cardiol 1998;32:275ñ282.
10. Vakili BA, Kaplan R, Brown DL. Volume-outcome relation for physicians and hospitals performing angioplasty for acute myocardial infarction in New York State. Circulation 2001;104:2171ñ2176.
11. Hosmer D, Lemeshow S. Applied Logistic Regression. New York, NY: John Wiley & Sons, 2000.
12. Srinivas VS, Brooks MM, Detre KM et al. Contemporary percutaneous coronary intervention versus balloon angioplasty for multivessel coronary artery disease: A comparison of the National Heart, Lung and Blood Institute Dynamic Registry and the Bypass Angioplasty Revascularization Investigation (BARI) study. Circulation 2002;106:1627ñ1633.
13. Singh M, Rihal CS et al. Comparison of Mayo Clinic risk score and American College of Cardiology/American Heart Association lesion classification in the prediction of adverse cardiovascular outcome following percutaneous coronary interventions. J Am Coll Cardiol 2004;44:357ñ361.
14. DeWood MA, Spores J, Notske R et al. Prevalence of total coronary occlusion during the early hours of transmural myocardial infarction. N Engl J Med 1980;303:897ñ902.
15. Myerburg RJ, Kessler KM, Castellanos A. Sudden cardiac death. Structure, function, and time-dependence of risk. Circulation 1992;85(Suppl I):I1ñI10.
16. Bigger JT Jr, Fleiss JL, Kleiger R et al. The relationships among ventricular arrhythmias, left ventricular dysfunction, and mortality in the 2 years after myocardial infarction. Circulation 1984;69:250ñ258.
17. Norris RM, Brandt PW, et al. Prognosis after recovery from first acute myocardial infarction: Determinants of reinfarction and sudden death. Am J Cardiol 1984;53:408ñ413.
18. Shirani J, Roberts WC. Comparison of modes of death and cardiac necropsy findings in fatal acute myocardial infarction in men and women > 75 years of age. Am J Cardiol 2000;86:1010ñ2,A8, A10.
19. Shirani J, Roberts WC. Major cardiac findings at necropsy in 366 American octogenarians. Am J Cardiol 1995;75:151ñ156.
20. Hochman JS, Sleeper LA, Webb JG, et al. Early revascularization in acute myocardial infarction complicated by cardiogenic shock. SHOCK Investigators. Should we emergently revascularize occluded coronaries for cardiogenic shock. N Engl J Med 1999;341:625ñ634.