Compliance with guidelines and 1-year mortality in patients with acute myocardial infarction: a prospective study

Transcription

1 Publish-Ahead-of-Print published January 28, 2005 European Heart Journal doi: /eurheartj/ehi107 Clinical research Compliance with guidelines and 1-year mortality in patients with acute myocardial infarction: a prospective study François Schiele*, Nicolas Meneveau, Marie France Seronde, Fiona Caulfield, Renaud Fouche, Gerard Lassabe, Denis Baborier, Pierre Legalery, and Jean-Pierre Bassand on behalf of the Réseau de Cardiologie de Franche Comté group Department of Cardiology, University Hospital Jean-Minjoz, Baulevard Fleming, Besançon, France Received 29 July 2004; revised 2 December 2004; accepted 9 December 2004 KEYWORDS Guidelines; Myocardial infarction; Risk score; Registry Aims In patients with acute myocardial infarction (MI), mortality can be predicted by risk scoring systems, but the impact of therapy recommended by guidelines is poorly documented. The aim of this study was to determine, taking into account the patient s condition at admission, to what extent the degree of guideline compliance influences the 1-year survival of patients admitted for acute MI. Methods and results A 6-month registry was carried out in a geographically limited area, prospectively including all patients with acute MI. A risk score based on initial presentation, and a compliance index based on patient characteristics, type of MI, in-hospital management (including revascularization strategies and use of recommended drugs) were established. Patients were clinically followed at 1 year. A total of 754 patients, 333 ST elevation MI and 421 non-st elevation MI, were included. The median compliance index (percentage of optimal compliance with guidelines) was 0.66 (95% CI 0.5;8.3). One-year mortality rate was 11.5%. By logistic regression, three variables were independently related to mortality: type of MI [OR ¼ 2.6 (1.5;4.3)], risk score [OR ¼ 2.4 (1.9;3.1) per additional 10%], and compliance index [OR ¼ 0.8 (0.7;0.9) per additional 10%]. Conclusion A clear relationship between the extent of guideline implementation, and 1-year mortality was shown and this relationship remained strong after stratification on the risk score at admission and the type of MI. These data emphasize the need for thorough implementation of guidelines to improve the outcome of patients suffering from acute MI. Introduction Guidelines for the management of patients with acute myocardial infarction (MI), with 1,2 and without 3 5 persistent ST segment elevation (STEMI or NSTEMI), provide * Corresponding author. Tel: þ ; fax: þ address: francois.schiele@ufc-chu.univ-fcomte.fr recommendations for acute care and best clinical practice. In the field of acute coronary syndromes, several strategies and drugs are recommended by guidelines on the basis of the demonstration of a reduction of mortality. Therefore, if widely applied in routine practice, strict application of recommendations should theoretically lead to better clinical outcome, namely a reduction in mortality. 6 & The European Society of Cardiology All rights reserved. For Permissions, please journals.permissions@oupjournals.org

2 Page 2 of 8 F. Schiele et al. However, large multinational registries that reflect the real life of patient management and outcome have shown that the uptake of guidelines into routine practice is suboptimal, and that there are significant geographical and temporal variations There is still a noteworthy underuse of reperfusion in acute STEMI, 11 and of invasive strategies or glycoprotein (GP) IIb/IIIa inhibitors in highrisk, non-st segment elevation acute coronary syndromes (ACSs), 8,12 as well as a deficit in the prescription of statins, anti-platelet drugs, or ACE-inhibitors. 13 Factors that may influence adherence to guidelines have been debated and are not fully understood. 7,11,12,14,15 Two ongoing projects are assessing the impact of the application of guidelines on clinical outcome in patients with NSTEMI 16 and with STEMI. 17,18 Moreover, since risk scores are strong predictors of early mortality in patients with NSTEMI, 10 STEMI, 9 or both types of MI, 19,20 these scores should be considered both in the degree of application of guidelines and in the assessment of clinical outcome. The aim of this study was to determine to what extent the degree of guideline compliance influences the 1-year mortality of patients admitted for acute MI, taking into account the patient s condition on admission. Methods Participating centres The study was designed and conducted by the Réseau Franc Comtois de Cardiologie and aimed at recording all admissions of patients suffering from MI. This group represents a network of all 12 cardiology centres in the region of Franche-Comté (a region in Eastern France with a population of 1.2 million inhabitants) where patients with chest pain or suspected ACS are admitted. Algorithms for patient management, based on the guidelines available at the time of the study design 2,3,5 were established and accepted by the physicians involved in the network. Within the region, medical transfers to one of the three centres with on-site coronary angiography and angioplasty facilities are available at any time. All patients gave informed consent and were included, leading to a cohort made up of all patients admitted for ACS in the region over the study period. Study design, population, and sample size Enrolment was performed over a 6-month period (from October 2000 to March 2001). All patients with a confirmed diagnosis of acute MI, with or without ST segment elevation according to the new definition of MI, 21 were included. At the time of study design, no data were available in our region about the number of patients admitted with acute MI, their risk level, or the management of these patients in the participating centres. One-year mortality was assumed to be close to that observed in other registries, at around 10%. We hypothesized that, in order to be able to perform a multivariable analysis including the compliance index and the risk level, a minimum of 50 events was needed. Thus, with an expected 10% 1-year mortality, at least 500 patients were needed. When, at 6 months, more than 500 patients had been included, we decided to stop inclusions. The study was exclusively supported by a research association (Association Franc Comtoise d Aide à la Recherche en Cardiologie), without direct support from any industry or administrative sources. Definitions and data collection Before the study commenced, standardized definitions were established and all the participating centres were informed through meetings and written notification. A dedicated team of data managers was available in the co-ordinating centre to assist with completion and verification of the five-page case report form. In total, 246 items were recorded, including demographic data, previous medication and diseases, clinical presentation, time interval between onset of symptoms and admission, treatment modalities (in-hospital and at discharge), transfer, coronary angiography or revascularization, in-hospital outcome, and rehabilitation. Patients were contacted at 1 year through telephone contact or a scheduled consultation to assess the efficacy endpoint, which was 1-year survival (all causes of death were considered). Computerized checks were performed to verify the coherence of the data, queries were generated in the case of inconsistencies, and a sample of medical records was reviewed in each centre. Risk stratification risk index We used a risk score derived from the Global Registry of Acute Coronary Events (GRACE) score. 19 This score was determined and validated for all types of ACS, taking into account the medical history (age, history of congestive heart failure or MI), findings at initial presentation (resting heart rate, systolic blood pressure, and presence of ST segment depression), and findings during hospitalization (level of serum creatinine and troponin release). Compliance with guidelines: compliance index The compliance score was the ratio of actual treatment to theoretical treatment. The theoretical treatment score was calculated for every patient, taking into account the type of MI, the treatment eligibility criteria, and the existence of contraindications to drugs or treatments. A rating of 0 was used in the case of specified contraindications, otherwise the score was calculated as follows: (i) STEMI: we attributed one point for the use of reperfusion therapy (primary or rescue percutaneous coronary intervention or thrombolysis) in patients admitted within the first 12 h after onset of symptoms. One point each was given for early (within the first 24 h) prescription of aspirin (or clopidogrel in cases of contraindication to aspirin), betablockers (atenolol, acebutolol, metoprolol, propranolol or timolol), ACE-inhibitors (only in patients with heart failure and/or left ventricular dysfunction or the presence of diabetes plus one major risk factor), and statins (simvastatine or pravastatine). (ii) NSTEMI: we attributed one point each for: early use of GP IIb/IIIa inhibitors, early invasive strategy (coronary angiography and/or revascularization within 48 h), heparin [unfractionated or low molecular weight heparin (LMWH)], beta-blockers, aspirin (or clopidogrel), ACE-inhibitors (same indication as for STEMI), and statins. Possible contraindications were defined for all treatments:. Thrombolysis: stroke within the previous 6 months, previous intracranial bleeding, active bleeding, systolic blood pressure.180 mmhg.

3 Compliance with guidelines and mortality in AMI Page 3 of 8. Angioplasty: severe renal failure (creatinine clearance below 30 ml/min).. Beta-blockers: bradycardia,50 b.p.m., systolic blood pressure below 80 mmhg, sign of heart failure (Killip class.1), left ventricular ejection fraction (LVEF) below 0.35, atrioventricular (AV) block on ECG, history of chronic obstructive lung disease or peripheral artery disease.. ACE-inhibitors: severe renal failure with creatinine clearance,30 ml/min. Aspirin or clopidogrel: active gastric bleeding. GPIIb/IIIa inhibitors: thrombocytopenia, /dL. Statins: history of previous intolerance. Thus, the maximum theoretical score was 5 for patients with STEMI and 7 for patients with NSTEMI. An index of the level of compliance (compliance index) was then calculated as the ratio of points for actual management to the theoretical score for that patient. Statistical analysis Categorical variables are presented as number of cases and percentage, scores as median (interquartiles), continuous variables as mean + standard deviation. Comparisons were performed using x 2, the Student s t-test, and likelihood x 2 tests, as appropriate. All tests were two-sided, a P-value of,0.05 was considered significant. One-year survival was presented using Kaplan Meier curves. Associations between variables and mortality were assessed using stepwise multivariable logistic regression and expressed as odds ratios [OR (95% Wald confidence interval)]. Variables that were candidates for inclusion in the multivariable model were determined by univariate relation with mortality (P, 0.20 by Wald x 2 test) and only those with P, 0.05 were retained in the final model. Thus, compliance index, risk score, type of MI, and history of peripheral artery disease were candidates, as well as gender, and centre. Separate logistic regression models were developed for the two continuous variables (compliance index and risk score) to Table 1 Demographic data and risk factors of the population test for a linear relationship with mortality. To ensure the absence of bias that could alter the prognostic value of the degree of guidelines application resulting from logistic regression model building procedures, we performed an additional logistic regression, where all candidate variables were forced into the model. Reliability, the concordance between predicted and observed outcomes, was assessed by grouping all patients into deciles of predicted risk and graphically comparing the observed prevalence of mortality for each centile group in the whole population and in four randomly selected subgroups. The c-statistic (a measure of model discrimination, equivalent to the area under a receiver operating characteristic curve) and the Hosmer and Lemeshow test (measure of model calibration) were used to determine the performance of the model. Analyses were performed using SAS software, version 8 (SAS Institute Inc.). Results Data collection During the 6 months of recruitment, 754 patients were admitted with a final diagnosis of MI, 333 (44%) with STEMI, 421 (56%) with NSTEMI. The demographic characteristics, risk factors, and signs or symptoms at presentation are displayed in Table 1. The median risk score was 113 (85;133). Management of patients with STEMI Of the 333 patients with STEMI, 227 (68%) were eligible for reperfusion strategy, which was actually performed in 176 (78%) patients. In 111 (63%) it was performed by thrombolytics, in 38 (22%) by primary angioplasty, and Variable STEMI NSTEMI P n Male 235 (70.5) 303 (72) 0.29 Age a 66 (15) 70 (13) 0.02 History of diabetes mellitus 79 (24) 100 (24) 0.94 History of high blood pressure 139 (42) 236 (56) 0.01 History of hyperlipidaemia 155 (47) 196 (47) 0.94 Smoker 178 (53) 209 (50) 0.33 History of previous PCI 27 (8) 71 (17) 0.01 History of previous MI 39 (12) 71 (17) 0.01 History of renal failure 9 (3) 23 (5) 0.06 Aspirin before 67 (20) 155 (37) Clopidogrel before 9 (3) 20 (5) 0.67 Statin before 31 (9) 77 (18) 0.01 Beta-blocker before 52 (16) 106 (25) 0.01 ACE-inhibitors before 43 (13) 106 (25) 0.01 Heart rate on admission a 79 (24) 80 (19) 0.14 Systolic blood pressure on admission a 136 (24) 143 (27) NSTEMI with ST segment depression 145 (34) Serum glucose level, mg/dl a 1.68 (1.17) 1.41 (0.69) Serum creatinine level, mg/dl a 14.9 (11) 15.1 (11) 0.44 Modified GRACE score a 103 (32) 113 (29) a Continuous data expressed as mean (SD) PCI: percutaneous coronary intervention; Before: treatment given before admission; GRACE: Global Registry of Acute Coronary Events.

4 Page 4 of 8 F. Schiele et al. Table 2 Theoretical and actual management of patients with STEMI Eligible patients n (%) Actually treated n (%) Ratio actual/eligible Reperfusion 227 (68) 176 (53) 0.78 ACE-inhibitors (discharge) 225 (68) 151 (45) 0.67 Aspirin/clopidogrel 333 (100) 304 (91) 0.91 Statins (discharge) 330 (99) 206 (62) 0.62 Beta-blockers 156 (47) 129 (39) 0.83 ACE: angiotensin converting enzyme. Table 3 Theoretical and actual management of patients with NSTEMI Eligible patients n (%) Actually treated n (%) Ratio actual/eligible ACE-inhibitors (discharge) 272 (65) 149 (35) 0.53 Aspirin/clopidogrel 415 (99) 381 (90) 0.92 Statins (discharge) 407 (97) 255 (61) 0.63 Beta-blockers 262 (62) 159 (38) 0.61 Heparin 415 (99) 389 (92) 0.94 LMWH 392 (93) 259 (62) 0.66 GPIIb/IIIa inhibitors 404 (96) 125 (30) 0.31 Early invasive strategy 398 (95) 254 (60) 0.64 Revascularization with GPIIb/IIIa inhibitors 398 (95) 111 (26) 0.28 in 27(15%) by thrombolysis followed by rescue angioplasty. The ratio of treated to eligible was 0.78 (176/ 227) for reperfusion strategy, 0.83 (129/156) for betablockers, 0.67 (151/225) for ACE-inhibitors, 0.91 (304/ 333) for aspirin (or clopidogrel or both), and 0.62 (206/ 330) for statins (Table 2 ). The median compliance index in this group was 0.76 (0.55;0.90). Management of patients with NSTEMI The ratio of treated to eligible patients was 92% (381/ 415) for aspirin, 61% (159/262) for beta-blockers, 63% (255/407) for statins, and 94% (389/415) for heparin, including 66% for LMWH. GPIIb/IIIa inhibitors were given in only 31% (125/404) of cases (whereas 96% of the patients were actually eligible for this treatment). Sixty-four per cent (254/398) were submitted to early coronary angiography and 28% (111/398) to revascularization with GPIIb/IIIa inhibitors (Table 3 ). The median compliance index was 0.66 (0.5;8.3). Since compliance index in both groups was normalized by dividing by the maximum for that individual, this made it possible to combine the two populations (STEMI and NSTEMI). Thus, the median compliance index was 0.66 (0.50;0.83). Factors associated with 1-year mortality Clinical follow-up was complete for all patients. At 1 year, 87/754 (11.5%) patients had died, 46/333 (13.8%) were STEMI patients and 41/421 (9.7%) were NSTEMI patients. One-year survival probabilities were, respectively, and (P ¼ 0.07 by log rank test). As expected, by univariate analysis, the risk score was found to be a strong predictor of mortality: a 10% increase in risk score corresponded to a 2.6-fold higher mortality rate [OR ¼ 2.6 (2.1;3.3), c-statistic ¼ 0.79]. From the first to the fourth quartile of risk score, the 1-year mortality was 1/191(0.2%), 11/190 (5.8%), 25/ 180 (13.9%), and 50/193 (25.9%). Respective 1-year survival probabilities were: , , , and (Figure 1 ). Despite a wide range of mortality across centres ( %), there was no significant centre effect (heterogeneity test for mortality), but conversely, there was a centre effect for the degree of application of guidelines: compliance with guidelines ranged from 77.9% to 43.7% and the test for heterogeneity was significant [F-value ¼ 11.72, P, by ANOVA with Bonferroni correction (significant if P, 0.008)]. Similarly, a relationship was observed between the compliance index and 1-year mortality: a 10% increase in the compliance index corresponded to a decrease in mortality [OR ¼ 0.74 (0.67;0.81), c-statistic ¼ 0.71]. By ascending quartile of compliance index, the mortality rates were 41/164 (25.0%), 25/238 (10.5%), 15/205 (7.3%), and 6/147 (4.1%); respective 1-year survival probabilities were: , , , and (Figure 2 ). Risk score and compliance index were not independent. An inverse relationship was found between risk index and compliance index: patients at higher risk had a lower compliance index. By univariate analysis, five variables were related to 1-year mortality, with a P-value of,0.20 and were candidates for the multivariable analysis: type of MI, gender, presence of diabetes, risk score, and compliance index.

5 Compliance with guidelines and mortality in AMI Page 5 of 8 Figure 1 Kaplan Meier survival curves according to the quartiles of risk score. Black line, first quartile of risk score. Red line, second quartile of risk score. Green line, third quartile of risk score. Blue line, fourth quartile of risk score. Figure 2 Kaplan Meier survival curves according to the quartiles of compliance index. Blue line, fourth quartile of compliance index. Green line, third quartile of compliance index. Red line, second quartile of compliance index. Black line, first quartile of compliance index. By multivariable logistic regression, only three variables remained in the model and were independent predictors of mortality: (i) the risk score [OR ¼ 2.4 (1.9;3.1), per additional 10%], (ii) the compliance score [OR ¼ 0.8 (0.7;0.9) per 10% increase], and (iii) the type of MI [OR ¼ 2.6 (1.5;4.3) for STEMI vs. NSTEMI]. The relationship between 1-year mortality, quartiles of compliance index, and quartiles of risk score is illustrated in Figure 3. The logistic regression with all five candidate variables forced into the model provided similar results. The assessment of reliability of the model showed a satisfactory discriminant accuracy, with a c-statistic ¼ 0.82 and an

6 Page 6 of 8 F. Schiele et al. Figure 3 One-year mortality rate according to risk index on admission and compliance index. Mortality: 1-year mortality rate (all causes). Compliance: index of compliance to guidelines (1 to 4 ¼ first to fourth quartile). Risk index: derived from the GRACE risk score (1 to 4 ¼ first to fourth quartile). adequate model calibration, with Hosmer Lemeshow goodness of fit test ¼ The plot of predicted vs. observed mortality (%) is presented in Figure 4, as an aid to visualization of the reliability of the model. Transformation of both the compliance index and the risk score into categories did not improve the performance of the model. Discussion In this study, we observed a clear relationship between the extent to which clinical guidelines are implemented and 1-year mortality in patients admitted for acute MI. In addition, this result persisted when adjusted for the level of risk of the patients, assessed at admission through a risk score, as well as the type of MI. Consistency with data from registries A cohort survey was the only applicable design suitable for the purpose of this study. Great care was taken to minimize the bias traditionally associated with cohort or registry studies, by standardizing the definitions, informing all the investigators in detail, paying careful attention to data collection and verification, and by using quality controls, all of which was overseen by a dedicated team of data managers. 22 The design of the registry to include all patients admitted in all centres in a geographically limited area, and also the relatively short inclusion period were chosen to reduce the risk of methodological bias. Lastly, we can assume that there was no economic confounding factor since, in the French medical insurance system, neither the patients nor the physicians are directly concerned with the cost of care. The characteristics of this study s population were comparable to the data obtained from large, multinational population-based registries, for example as regards the proportion of risk factors, age and gender distribution, and type of MI (STEMI or NSTEMI). 7,23,24 Figure 4 Plot of predicted versus observed mortality (per cent) for visualization of the reliability of the model. The solid line represents perfect reliability. The population is divided into deciles of predicted mortality. The black dots represent observed mortality rate per decile in the whole population, the red and blue squares, the brown dots, and the yellow triangles represent observed mortality in randomly selected subgroups. The rate of reperfusion in patients with STEMI admitted less than 12 h after onset of symptoms is often considered a landmark figure. In our study, reperfusion therapy was performed in 77% of all eligible patients (53% of the whole population admitted with STEMI), figures comparable to those reported in previous published registries. 11,24 In NSTEMI patients, the use of GPIIb/IIIa inhibitors was the least frequently applied recommendation, prescribed in only 31% of this target population, findings which are again similar to previous publications. 12,25 At 1 year, the mortality rate was 11.5%; 13.8% in the group with STEMI vs. 9.7% in patients with NSTEMI, similar to that in the Euro Heart Survey of ACS. 24 Risk stratification Different risk scoring systems for patients with ACS have been published, usually based on the results of randomized studies and limited to patients with either NSTEMI 10 or STEMI. 9 In our study, the risk score was based on the GRACE score registry, which has been established in a registry population, validated for all types of ACS, 19 and therefore applicable in our population. Compliance with recommended treatments, risk level, and 1-year mortality In our study, we did not limit the assessment of compliance to the use of one particular drug or strategy, but rather we considered all the strategies recommended in the guidelines, associated with a reduction in mortality.

7 Compliance with guidelines and mortality in AMI Page 7 of 8 We did not try to assess which intervention was associated with the most substantial effect on mortality, but rather, we considered the global adherence to guidelines. Thus, we attributed the same weight to all the therapeutic measures considered (when applicable), namely reperfusion therapy, early invasive strategy, antiplatelet agents, heparin, beta-blockers, ACE-inhibitors, GPIIb/IIIa inhibitors, and statins. At the current time, no standardized method exists for assessing the degree of compliance with clinical guidelines, and variations exist in the quality indicators. In previous studies designed to assess compliance with guidelines, only ideal patients (i.e. those eligible for recommended treatment 26 or without contraindication to beta-blockers or aspirin 6 ) were included. In our study, since all patients admitted with acute MI were included, it was necessary to take into account all possible contraindications, including temporary contraindications. For example, we considered as contraindications to beta-blocker treatment, not only chronic obstructive lung or peripheral artery diseases, but also low blood pressure, signs of heart failure, bradycardia, or AV block that may have been transitory. As a result, only 47% of patients with STEMI and 62% with NSTEMI were considered eligible for beta-blockers. This rate is lower than in other registries, 27 but we wanted to avoid any overestimation of underuse of the recommended drug. In this study, we did not observe any centre effect, as described in other registries, 14 depending on the presence of on-site angioplasty facilities or the nature of the physicians caring for the patient. When the variable centre was forced into the multivariable analysis, it did not affect the odds ratio of the other variables; tests for model calibration and model discrimination were unchanged (Hosmer and Lemeshow goodness-of-fit test ¼ and c-statistic ¼ 0.824). We explain the lack of centre effect by the presence of the measure of compliance with guidelines in the model. Taking into account the level of risk, our results are consistent with previous reports that guidelines are applied less thoroughly in the patients at highest risk. 8,12 Nevertheless, our results cannot provide any explanation for this relationship. The impact of the degree of guideline application on mortality is an important result. This is in agreement with findings from two similar ongoing studies: CRUSADE (Can rapid risk stratification of unstable angina patients suppress adverse outcomes with early implementation of the ACC and AHA guidelines?) 16 where a fall of 11% in mortality for every 10% increase in adherence to ACC and AHA guidelines for patients with suspected MI was found; and GAP, 26,28 where the risk of death within a year was 25% lower after discharge if hospitals followed national guidelines and taught their patients how to follow those standards at home. However, since the risk level and the degree of compliance with guidelines are related, both variables must be assessed in a multivariable analysis, which, to date, has never been done. As a result, we observed that both risk score and compliance with guidelines are independent predictors of 1-year mortality. Thus, the higher mortality in patients with lower compliance with the guidelines cannot be solely explained by a higher risk level or by co-morbidities of these patients. Study limitations Despite the great attention paid, this study has several inherent limitations associated with cohort studies. The geographical and time-specific design may limit the extrapolation of the results to other medical centres and patients. In this study, only mortality was assessed and not occurrence of non-fatal recurrence of ACS, cardiac failure, or stroke. The assessment of the guideline compliance is debatable, notably the definition of contraindications for beta-blockers. Lastly, in this observational study, while we can conclude that patients with the most thorough application of the guidelines had a better outcome, we cannot address other important issues, such as: (i) is any improvement in the application of the guidelines possible or beneficial? (ii) which intervention, among those recommended in the guidelines, is the most beneficial? and (iii) why do patients at higher risk receive fewer effective treatments? Conclusions This study confirms the relationship between the degree of application of clinical guidelines and mortality rate in patients admitted for acute MI. It also shows, for the first time, that this relationship remains strong after stratification on the type of MI and baseline risk. We can speculate that greater uptake and more thorough application of clinical guidelines may reduce mortality from acute MI. Further investigations, using large multinational registries, are mandatory to check whether a more thorough application of the guidelines would result in a better outcome. Nonetheless, our results provide a strong argument in favour of continued educational efforts to increase the uptake of guidelines in clinical practice. References 1. Van de Werf F, Ardissino D, Betriu A, Cokkinos DV, Falk E, Fox KA et al. Management of acute myocardial infarction in patients presenting with ST-segment elevation. The Task Force on the Management of Acute Myocardial Infarction of the European Society of Cardiology. Eur Heart J 2003;24: Ryan TJ, Antman EM, Brooks NH, Califf RM, Hillis LD, Hiratzka LF et al update: ACC/AHA Guidelines for the Management of Patients With Acute Myocardial Infarction: Executive Summary and Recommendations: A report of the American College of Cardiology/ American Heart Association Task Force on Practice Guidelines (Committee on Management of Acute Myocardial Infarction). Circulation 1999;100: Bertrand ME, Simoons ML, Fox KA, Wallentin LC, Hamm CW, McFadden E et al. Management of acute coronary syndromes: acute coronary syndromes without persistent ST segment elevation; recommendations of the Task Force of the European Society of Cardiology. Eur Heart J 2000;21: Bertrand ME, Simoons ML, Fox KA, Wallentin LC, Hamm CW, McFadden E et al. Management of acute coronary syndromes in patients presenting without persistent ST-segment elevation. Eur Heart J 2002;23:

8 Page 8 of 8 F. Schiele et al. 5. Braunwald E, Antman EM, Beasley JW, Califf RM, Cheitlin MD, Hochman JS et al. ACC/AHA guidelines for the management of patients with unstable angina and non-st-segment elevation myocardial infarction: executive summary and recommendations. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on the Management of Patients with Unstable Angina). Circulation 2000;102: Mukherjee D, Fang J, Chetcuti S, Moscucci M, Kline-Rogers E, Eagle KA. Impact of combination evidence-based medical therapy on mortality in patients with acute coronary syndromes. Circulation 2004;109: Fox KA, Goodman SG, Anderson FA, Jr, Granger CB, Moscucci M, Flather MD et al. From guidelines to clinical practice: the impact of hospital and geographical characteristics on temporal trends in the management of acute coronary syndromes. The Global Registry of Acute Coronary Events (GRACE). Eur Heart J 2003;24: Yusuf S, Flather M, Pogue J, Hunt D, Varigos J, Piegas L et al. Variations between countries in invasive cardiac procedures and outcomes in patients with suspected unstable angina or myocardial infarction without initial ST elevation. OASIS (Organisation to Assess Strategies for Ischaemic Syndromes) Registry Investigators. Lancet 1998;352: Morrow DA, Antman EM, Charlesworth A, Cairns R, Murphy SA, de Lemos JA et al. TIMI risk score for ST-elevation myocardial infarction: A convenient, bedside, clinical score for risk assessment at presentation. An intravenous npa for treatment of infarcting myocardium early II trial substudy. Circulation 2000;102: Antman EM, Cohen M, Bernink PJ, McCabe CH, Horacek T, Papuchis G et al. The TIMI risk score for unstable angina/non-st elevation MI: A method for prognostication and therapeutic decision making. JAMA 2000;284: Eagle KA, Goodman SG, Avezum A, Budaj A, Sullivan CM, Lopez- Sendon J. Practice variation and missed opportunities for reperfusion in ST-segment-elevation myocardial infarction: findings from the Global Registry of Acute Coronary Events (GRACE). Lancet 2002;359: Peterson ED, Pollack CV, Jr., Roe MT, Parsons LS, Littrell KA, Canto JG et al. Early use of glycoprotein IIb/IIIa inhibitors in non-st-elevation acute myocardial infarction: observations from the National Registry of Myocardial Infarction 4. J Am Coll Cardiol 2003;42: Gandjour A, Kleinschmit F, Lauterbach KW. European comparison of costs and quality in the treatment of acute myocardial infarction ( ). Eur Heart J 2002;23: Gottwik M, Zahn R, Schiele R, Schneider S, Gitt AK, Fraunberger L et al. Differences in treatment and outcome of patients with acute myocardial infarction admitted to hospitals with compared to without departments of cardiology; results from the pooled data of the Maximal Individual Therapy in Acute Myocardial Infarction (MITRA 1 þ 2) Registries and the Myocardial Infarction Registry (MIR). Eur Heart J 2001;22: Scirica BM, Moliterno DJ, Every NR, Anderson HV, Aguirre FV, Granger CB et al. Differences between men and women in the management of unstable angina pectoris (The GUARANTEE Registry). The GUARANTEE Investigators. Am J Cardiol 1999;84: Peterson E, Roe MT, Lytle B, Newby K, Fraulo E, Gibler WB et al. The association between Care and Outcomes in Patients with Acute Coronary Sydrome: National Results From CRUSADE. JAm Coll Cardiol 2004;43(Suppl. A): Eagle KA, Gallogly M, Mehta RH, Baker PL, Blount A, Freundl M et al. Taking the national guideline for care of acute myocardial infarction to the bedside: developing the guideline applied in practice (GAP) initiative in Southeast Michigan. Jt Comm J Qual Improv 2002; 28: Mehta RH, Montoye CK, Gallogly M, Baker P, Blount A, Faul J et al. Improving quality of care for acute myocardial infarction: The Guidelines Applied in Practice (GAP) Initiative. JAMA 2002;287: Granger CB, Goldberg RJ, Dabbous O, Pieper KS, Eagle KA, Cannon CP et al. Predictors of hospital mortality in the global registry of acute coronary events. Arch Intern Med 2003;163: Eagle KA, Lim MJ, Dabbous OH, Pieper KS, Goldberg RJ, Van de Werf F et al. A validated prediction model for all forms of acute coronary syndrome: estimating the risk of 6-month postdischarge death in an international registry. JAMA 2004;291: Myocardial infarction redefined a consensus document of The Joint European Society of Cardiology/American College of Cardiology Committee for the redefinition of myocardial infarction. J Am Coll Cardiol 2000;36: Alpert JP. Are data from clinical registries ofany value? Eur Heart J 2000;21: Fox KA, Cokkinos DV, Deckers J, Keil U, Maggioni A, Steg G. The ENACT study: a pan-european survey of acute coronary syndromes. European Network for Acute Coronary Treatment. Eur Heart J 2000;21: Hasdai D, Behar S, Wallentin L, Danchin N, Gitt AK, Boersma E et al. A prospective survey of the characteristics, treatments and outcomes of patients with acute coronary syndromes in Europe and the Mediterranean basin; the Euro Heart Survey of Acute Coronary Syndromes (Euro Heart Survey ACS). Eur Heart J 2002;23: Fox K, Goodman SG, Klein W, Brieger D, Steg G, Dabbous O et al. Management of acute coronary syndroms. Variations in practice and outcome. Findings from the Global Registry of Acute Coronary Events (GRACE). Eur Heart J 2002; Mehta RH, Montoye CK, Gallogly M, Baker P, Blount A, Faul J et al. Improving quality of care for acute myocardial infarction: The Guidelines Applied in Practice (GAP) Initiative. JAMA 2002;287: Steg PG, Goldberg RJ, Gore JM, Fox KA, Eagle KA, Flather MD et al. Baseline characteristics, management practices, and in-hospital outcomes of patients hospitalized with acute coronary syndromes in the Global Registry of Acute Coronary Events (GRACE). Am J Cardiol 2002;90: Riba AA, Baker P, Roychoudhury C, Chen A, Mehta RH, Eagle KA et al. Guideline-based standardized care substantially reduces mortality in Medicare Patients with acute myocardial infarction: the American College of Cardiology s Guidelines Applied in Practice Program in Michigan. JACC 2004;43(Suppl. A):