Methods The PARAGON-B trial

Transcription

1 Misreporting of myocardial infarction end points: Results of adjudication by a central clinical events committee in the PARAGON-B trial Kenneth W. Mahaffey, MD, a Matthew T. Roe, MD, a Christopher K. Dyke, MD, a L. Kristin Newby, MD, a Neal S. Kleiman, MD, b Patty Connolly, RN, a Lisa G. Berdan, PA-C, MHS, a Rodney Sparapani, MS, a Kerry L. Lee, PhD, a Paul W. Armstrong, MD, c Eric J. Topol, MD, d Robert M. Califf, MD, a and Robert A. Harrington, MD, a for the PARAGON-B Investigators Durham, NC, Edmonton, Alberta, Canada, and Cleveland, Ohio Background Myocardial (re)infarction (MI), a common trial end point, can be difficult to identify because of inconclusive signs and symptoms. We examined disagreement between investigator and clinical events committee (CEC) reporting of MIs in an international, randomized trial. Methods The primary end point of the PARAGON-B trial was a 30-day composite of death, MI (CEC adjudicated), or ischemia-driven intervention. If CEC and investigator determinations of MI differed, we sent investigators event summaries and rationales for CEC decisions and asked whether they now agreed with the CEC assessment. If they still disagreed, they were to provide a rationale and supporting data. Such cases were reviewed, and a final decision was made. Results Overall, 1736 of 5225 (33%) patients had suspected MIs; the CEC adjudicated 483 of 1736 (28%) as MIs. In 404 patients (23%), investigator and CEC assessments of MI differed; 270 MIs were identified by the CEC but not investigators, and 134 were identified by investigators but not the CEC. Most disagreements concerned periprocedural MIs, but some reflected clinical ischemia and enzyme elevations. Letters for 382 disagreements were sent and returned by investigators, and investigators came to agree with CEC assessments in 307 cases (80%). For the other 75 cases (20%), after review the investigators assessments were confirmed in 10 cases, and the original CEC decisions were supported in the other 65 cases. Conclusions Investigators misreport MI end points, but most later agree with CEC assessments. These data support standard, independent adjudication of suspected MIs for accurate reporting, which may affect evaluations of therapies, sample-size calculations, and event-rate comparisons across trials. (Am Heart J 2002;143:242-8.) Myocardial infarction (MI) is an important nonfatal event in patients with acute coronary syndromes. In recent trials of new antiplatelet and antithrombin therapies in this population, prevention of (re)infarction has been a key component of composite clinical outcomes. 1-5 Determining whether a patient has had an MI can sometimes be difficult in clinical practice, however, because of unclear symptoms, inconsistent enzyme data, inconclusive electrocardiographic (ECG) changes, and cardiac revascularization procedures. Historically, From the a Duke Clinical Research Institute, Durham, NC, the b Baylor College of Medicine and Methodist Hospital, Houston, Tex, the c University of Alberta, Edmonton, Alberta, Canada, and the d Cleveland Clinic Foundation, Cleveland, Ohio. Supported by a grant from Hoffman-La Roche, Ltd, Basel, Switzerland. Guest Editor for this manuscript was Christopher P. Cannon, MD, Brigham and Women s Hospital, Boston, Mass. Submitted March 15, 2001; accepted June 14, Reprint requests: Kenneth W. Mahaffey, MD, Duke Clinical Research Institute, PO Box 17969, Durham, NC mahaf002@mc.duke.edu. Copyright 2002 by Mosby, Inc /2002/$ /1/ doi: /mhj the standard World Health Organization (WHO) definition of MI requires 2 of 3 characteristics (typical symptoms, enzyme rise and fall, and typical ECG pattern). This definition is outdated because current clinical practice is markedly different and the importance of enzyme elevations after revascularization procedures is now better appreciated. The use of clinical events committees (CECs) is widely accepted to adjudicate suspected nonfatal end point events in various cardiovascular trials, but only limited information has been published about such adjudication in cardiovascular investigations Our rationale for CEC adjudication is the need for systematic, unbiased, independent, and standard assessment of MIs because MI is an important nonfatal end point in the evaluation of new therapies in acute coronary syndromes. In a prior retrospective analysis, we reported the reasons for disagreements between the site-investigator reported MIs and the CEC adjudication of MI in the Platelet Glycoprotein IIb/IIIa in Unstable Angina: Receptor Suppression Using Integrilin (Eptifibatide) Therapy (PURSUIT) trial. 13 To better understand the

2 American Heart Journal Volume 143, Number 2 Mahaffey et al 243 basis for these disagreements, we prospectively planned this analysis, a study of the identification of MIs in the second Platelet IIb/IIIa Antagonist for the Reduction of Acute Coronary Syndrome Events in a Global Organization Network Trial (PARAGON-B). Methods The PARAGON-B trial The PARAGON-B trial examined the role of lamifiban, an intravenous platelet glycoprotein IIb/IIIa antagonist, in patients with acute coronary syndromes without persistent ST-segment elevation. The trial enrolled 5225 patients in 29 countries and used the previously reported inclusion and exclusion criteria and treatment regimens. 15 The primary end point was a composite of death, MI, or severe recurrent ischemia leading to urgent or unplanned intervention by 30 days. The CEC adjudicated all suspected recurrent ischemic and MI end points, and the adjudicated data were used in the primary efficacy analyses. Definitions The protocol defined MI as an end point on the basis of clinical, ECG, and laboratory criteria (Appendix). Mycardial infarctions that occurred before or at enrollment were not included in the primary end point. The site investigators and the CEC used the same MI criteria. These criteria were presented in the protocol, at investigator meetings, and in trial materials and newsletters. Data collection Data for all patients were collected on standard case report forms. Information collected included data on all cardiac enzymes, ECGs during the baseline hospitalization and at 30 days, revascularization procedures, detail of ischemic episodes, clinical complications, medications, and readmission records. All enzyme values for each patient during the baseline hospitalization were recorded, and study monitors at the coordinating centers verified them against source documents from the site. The protocol mandated serial assessment of cardiac enzyme after all recurrent ischemic events and revascularization procedures. Some countries in eastern and western Europe and Latin America (~700 of all patients enrolled) were asked to send blood samples to a central core laboratory for creatine kinase (CK) and CK-MB analyses in addition to performing enzyme analyses at their own institutions. In prior studies, laboratory inconsistencies had been observed in centers from these countries that made MI adjudication more difficult. An independent, blinded, ECG core laboratory read all ECGs and identified suspected MIs. Clinical events classification process The structure of the CEC and the event-adjudication process used in the PARAGON-B trial were similar to those used in our group s prior studies. 16 In brief, computer algorithms systematically identified key clinical, cardiac enzyme, and ECG data from the database that could indicate the occurrence of an MI. These variables were chosen for broad identification of all possible MIs. For each patient with a suspected MI identified by the algorithm, a case folder was prepared with all available clinical information (worksheet containing key case report form data, all cardiac enzymes, all ECGs, and available discharge summaries) and was reviewed independently by 2 physicians blinded to treatment. To maintain consistency in the reviews, we used a group of only 10 physician reviewers. If the 2 physicians agreed that an MI had or had not occurred, the case was classified as resolved. A committee of faculty cardiologists reviewed and adjudicated by consensus those cases about which the 2 CEC physicians disagreed. Reconciliation of disagreements between the site investigators and CEC After the CEC had completed adjudication of each suspected MI event, it was determined whether disagreement existed between the CEC and the site investigator regarding the occurrence of MI. If there was disagreement, a letter was sent to the site investigator. The letter provided a brief clinical summary of the event adjudicated by the CEC and provided the rationale for the CEC decision. The investigator was asked to tick 1 of 2 boxes on the letter signifying agreement or disagreement with the CEC assessment. If investigators disagreed with the CEC assessment, they were asked to provide a rationale for their disagreement. Additional cardiac enzyme data, ECGs, or medical records could be submitted to support the site investigators decisions. In cases in which the investigator continued to disagree with the CEC (once the reconciliation process was initiated), the suspected event was reviewed by a committee of faculty cardiologists (K. A. M., M. T. R., R. A. H., L. A. N.), and a final decision was reached by consensus. If necessary, the individual site investigator was contacted by telephone to discuss the event (~10% of cases). Statistical analysis Variables were summarized as percentages for dichotomous variables or as medians (25th and 75th percentiles) for continuous variables. Efficacy analyses were based on logistic regression analysis that adjusted for baseline variables known to be predictors of outcomes. 5 Results The PARAGON-B trial enrolled 5225 patients. Table I shows the number of patients with suspected events identified and adjudicated by the CEC. Overall, 1736 of 5225 (33%) patients were identified with a suspected MI event, and 483 of 1736 (28%) of these events were classified as MI by the CEC according to the protocol MI definitions. The CEC identified more MIs than site investigators reported in each region. Disagreement between the site-investigator assessment and the CEC assessment of MI occurred in 404 of 1736 (23%) patients with suspected MI (Table II). Of these 404 patients with disagreements, 270 had an end point MI adjudicated by the CEC but not reported by the site investigator, and 134 patients had MI reported by the site investigator but not adjudicated as an end

3 244 Mahaffey et al American Heart Journal February 2002 Table I. Patients with suspected MI Enrolled Suspected MI CEC-adjudicated MI Site-reported MI Region n n (%)* n (%) n (%) Asia-Pacific (35) 51 (10) 43 (9) Europe (32) 228 (8) 167 (6) South America (27) 28 (13) 19 (9) North America (35) 176 (10) 118 (7) Overall (33) 483 (9) 347 (7) *Percentage of patients enrolled. Table II. Disagreements between site-investigator reporting and CEC assessment of MI before reconciliation with the site Suspected MI Disagreement CEC Yes/Site No CEC No/Site Yes Region n (%)* n n n Asia-Pacific 174 (35) Europe 877 (32) South America 56 (27) North America 629 (35) Overall 1736 (33) *Percentage of patients enrolled per region. Figure 1 Table III. Types of cases with disagreements between site investigator and CEC before reconciliation CEC Yes/ CEC No/ Overall Site No Site Yes Type of case n n n Related to clinically evident ischemia Related to percutaneous intervention Related to bypass surgery Related to death* Elevated CK-MB only ECG changes only MI reported by investigator and not CEC Unknown type Overall Agreements and disagreements between CEC and site investigators. Percent based on all patients (N = 5225). *Clinically significant cardiac event resulting in death (versus sudden death without evidence of infarction). point MI by the CEC. Figure 1 shows the number of cases with agreement and disagreement between the CEC and site-investigator assessments of MI in the entire trial. The proportion of patients with disagreement between the CEC and the site investigators was similar across geographic regions. Table III provides data on the type of MIs for which the CEC and the site investigators had disagreements before reconciliation. Most of the events were infarctions around the time of revascularization procedures, although a number of cases were associated with clinical evidence of ischemia and enzyme elevations. The reconciliation of disagreements between the siteinvestigator and CEC assessments of MI is shown in Figure 2 and Table IV. A total of 382 letters were sent to and returned from site investigators. No letters were sent for 22 cases with disagreements for the following reasons: in 2 of these cases, the site investigator reported an MI that clearly had occurred more than 24 hours before randomization and thus was not a postenrollment MI;

4 American Heart Journal Volume 143, Number 2 Mahaffey et al 245 Figure 2 Figure 3 Reconciliation of disagreements between site-investigator and CEC assessments of MI. *Twenty patients had core laboratory data identifying MI, and 2 patients had MI >24 hours before enrollment. Reconciliation of disagreements in which the CEC identified postenrollment MI and site investigator did not. *Eighteen patients had core laboratory data identifying MI. Table IV. Types of cases with disagreements between site investigator and CEC after reconciliation CEC Yes/ CEC No/ Overall Site No Site Yes Type of case n n n Related to clinically evident ischemia Related to percutaneous intervention Related to bypass surgery Related to death* Elevated CK-MB only ECG changes only Enrollment MI (not post enrollment) Unknown (site-reported MI only) Overall *Clinically significant cardiac event resulting in death (vs sudden death without evidence of infarction). and in 20 cases, the CEC identified MI on the basis of review of the core-laboratory enzyme data, to which the site investigators did not have access. Of the 382 letters sent, the site investigator agreed with the CEC assessment in 307 cases (80%). In 75 (20%) of the disagreements, the site investigator continued to disagree with the CEC assessment and did not change the original report. A faculty committee of cardiologists reviewed these 75 cases. In 10 of these cases, the committee did not agree with the original CEC decision and supported the investigator assessment. In the other 65 cases, the committee confirmed the original CEC decision. Figure 3 shows reconciliation of disagreements in which the CEC identified a postenrollment MI and the site investigator did not, and Figure 4 shows the converse (the CEC did not identify a postenrollment MI, Figure 4 Reconciliation of disagreements in which CEC did not identify postenrollment MI and site investigator did. *Two patients had core laboratory data without evidence of MI, and 2 patients had MI before enrollment. and the site investigator did). The number of cases in which the investigator later agreed with the original CEC decision was similar, despite the type of disagreement. The proportion of cases in which the site investigator continued to disagree with the CEC decision, as detailed on the reconciliation letter, was also similar. In PARAGON-B, lamifiban resulted in a nonsignificant 9% relative reduction in the primary composite end point. We reconstructed the primary efficacy analysis by using the site-investigator assessments before reconciliation (Table V). The absolute and relative treatment effects were similar whether the investigator-reported or CEC-adjudicated MI data were used. Discussion Determination of MI can sometimes be difficult because of conflicting patient history, clinical symp-

5 246 Mahaffey et al American Heart Journal February 2002 Table V. Primary efficacy analysis with investigator-reported versus CEC-adjudicated data Overall Placebo Lamifiban Event (n = 5163) (n = 2568) (n = 2595) P value* Death, infarction, or recurrent ischemia CEC 12.3% 12.8% 11.8%.33 Investigator 9.4% 9.7% 9.1%.57 Death or infarction CEC 11.0% 11.5% 10.6%.32 Investigator 8.6% 8.9% 8.2%.42 *By logistic regression. toms, cardiac enzyme data, and ECG information. In the PARAGON-B trial, disagreement after CEC identification of MI was not uncommon. Site investigators underreported MI events, and, with prospective reconciliation of disagreements between the site investigators and the CEC, the site investigators agreed with the CEC assessment of MI in most cases. These results are consistent with findings from trials in which a similar CEC group adjudicated suspected MIs in comparable patient populations. 9,13 The CEC identified 270 MIs that were not reported by site investigators. After reconciliation with site investigators, 77% of these cases were confirmed as MIs by the investigators. Conversely, 134 events were reported as MIs by the investigators. These events did not meet study criteria for MI, however, and during reconciliation, site investigators agreed with the CEC in 86% of these cases. The reason for misreporting of MIs by site investigators is not entirely clear. The definitions for MI in PARAGON-B were developed by an international steering committee on the basis of their experience and clinical expertise. Because of the broad geographic enrollment planned for PARAGON-B, definitions were designed to be applicable for an array of clinical practice situations. The definitions were detailed in the study protocol and in study materials, so that the CEC and site investigators had the same set of criteria by which to define MIs. Most cases of disagreement were events that occurred around the time of percutaneous or surgical revascularization procedures. These are the same types of MIs commonly misclassified by investigators in our prior analyses. 9,13 This finding is also consistent with an ongoing controversy in clinical practice in which some physicians are reluctant to classify enzyme elevations after revascularization procedures as MI although such elevations are clearly associated with worse clinical outcomes. 6 The methods of data collection may have contributed to the misreporting of MI. MI is often a complex clinical entity, and the use of standard data collection forms as used in the PARAGON-B trial may have contributed to inaccuracies in reporting. In addition, although the collection of MI end points on the form was made with simple yes or no check-boxes for MI, not all MIs are the same. Investigators may have reported only those that manifested themselves with the classic presentation of chest pain, ECG changes, and positive cardiac markers. Further refinement of the forms may be helpful, and changes could prompt investigators to be more comprehensive in MI reporting. Rationale for CEC These results support the need for independent identification and adjudication of MI end points in clinical investigation. Investigators underreport MI end points when standard case report forms are used. The accurate assessment of event rates is important in clinical investigation for several reasons. First, accurate assessment of MI rates in both experimental groups is essential to detect true treatment differences. Although physicians are typically blinded to treatment, if a treatment preferentially reduced a certain type of MI (such as those after percutaneous intervention) and these MIs were systematically not reported by investigators, the ability to detect a treatment benefit could be compromised without CEC adjudication. Second, event rates are critically important in sample-size and power calculations. Reliance on incorrect information might result in underpowered studies of promising therapies in similar patient populations. Third, event rates in clinical trials are used not only to analyze treatment effects but also to understand the morbidity and mortality in patients with the disease being studied. Fourth, comparison of MI event rates between trials of similar populations is already confounded by several issues, including MI definitions. Such comparisons would be compromised further if they needed to account for CEC-determined versus investigator-determined events. The strategy used by a CEC to adjudicate MIs can dramatically influence event rates and the proportion of events with disagreements between site investigators and a CEC. Some trials have confirmed only events reported by the investigators, 7-11,17 whereas others have adjudicated all suspected events identified by systematic screening of patient data. 1-3,12,18,19 When only

6 American Heart Journal Volume 143, Number 2 Mahaffey et al 247 investigator-reported events are reviewed by a CEC, the reported event rates will be identical to or lower than the site-investigator rates. When events are identified independently by a CEC, the CEC event rates may be higher, lower, or the same as the site investigator reported rates. Effect of CEC on treatment effect The absolute reduction in the composite incidence of death or MI with lamifiban treatment was comparable when either CEC-adjudicated data or investigatorreported data (0.9% vs 0.7%) was used. Some studies have shown larger differences in the observed treatment effect when CEC data are used instead of site investigator reported MIs, 1,3,16 but others have not. 7-9 The reason for an enhanced treatment effect in some trials with CECadjudicated data but not in others is unclear. Implications These convincing data add further support for the belief that CEC adjudication of suspected nonfatal MI end point events is important for independent, unbiased, standard, systematic assessments in clinical trials. Furthermore, prospective efforts to reconcile disagreements between the site investigator and the CEC may not be necessary. Although such a reconciliation process provides added confidence in the end point assessment, it is a laborious effort requiring substantial time and monetary resources. In the current study, the original CEC decision was changed in only 10 cases, which represented 0.6% of the 1736 cases reviewed by the CEC or 2.6% of the 382 cases reconciled directly with the investigator, despite the comprehensive reconciliation process with site investigators. Education of site investigators regarding the importance of accurate event reporting may help reduce disagreements between the site investigators and the CEC. In addition, review of MI criteria with investigators may enhance the reporting of this end point. Clinical trialists can help by developing data collection forms that are clear and easy to complete. Finally, standard definitions for MI are needed in both clinical trials and clinical practice. Consensus about the criteria for MI in clinical trials, particularly if they mimic those in clinical practice, may help avoid confusion. Several recent large clinical trials have used different definitions for MI. 1,3-5 A recent consensus statement by the American College of Cardiology and the European Society of Cardiology redefining the MI is an important advance to make definitions more contemporary and model current clinical practice. 20 Conclusion Myocardial infarction is an important clinical outcome, and its reduction is a key efficacy measure of new therapies being evaluated in patients with acute coronary syndromes. Site investigators in the PARAGON-B trial did not accurately report MI end points with standard case report form tools. A prospective, interactive reconciliation of disagreements between the site investigators and the CEC showed that investigators agreed with the CEC assessment of MI in an overwhelming majority of cases. These data support the need for standardized, independent, CEC adjudication of suspected MIs. Accurate assessment of event rates has important implications in clinical trials for the comprehensive evaluation of new therapies, proper sample size calculations, and the comparison of event rates between different clinical trials. We thank Drs John H. Alexander, Cecilia M. Bahit, Mauricio G. Cohen, Mark A. East, Adam B. Greenbaum, Michael P. Hudson, David F. Kong, David O. Martin, Darren K. McGuire, J. Conor O Shea, Monica R. Shah, and Eric J. Velazquez, and Freda Wood, RN, for their efforts in event adjudication. We thank Pat French for editorial assistance. References 1. GUSTO-IIb Investigators. A comparison of recombinant hirudin with heparin for the treatment of acute coronary syndromes. N Engl J Med 1996;335: PRISM Investigators. A comparison of aspirin plus tirofiban with aspirin plus heparin for unstable angina. N Engl J Med 1998;338: PRISM PLUS Investigators. Inhibition of the platelet glycoprotein IIb/IIIa receptor with tirofiban in unstable angina and non-q-wave myocardial infarction. N Engl J Med 1998;338: PURSUIT Investigators. Inhibition of platelet glycoprotein IIb/IIIa with eptifibatide in patients with acute coronary syndromes without persistent ST-segment elevation: a randomized, placebo-controlled, clinical trial. N Engl J Med 1998;339: PARAGON-B Investigators. Results of an international, randomized, controlled trial of lamifiban, heparin, or both in unstable angina. Circulation In press. 6. Califf RM, Abdelmeguid AE, Kuntz R, et al. Myonecrosis after revascularization procedures. J Am Coll Cardiol 1998;31: Mahaffey KW, Granger CB, Woodlief L, et al. Centralized systematic adjudication of clinical endpoints in multicenter trials of acute coronary syndromes identifies patients at high risk for adverse clinical outcomes [abstract]. J Am Coll Cardiol 1996;27:250A. 8. Tardiff BE, Mahaffey KW, Sigmon KN, et al. Differences in endpoint identification by investigators versus independent review [abstract]. Control Clin Trials 1996;17:25S. 9. Mahaffey KW, Tardiff BE, Granger CB, et al. Comparison of investigator and central events review of myocardial infarction rates as a means to improve events classification: results from GUSTO-IIa [abstract]. Control Clin Trials 1996;17:136S. 10. Ives DG, Fitzpatrick AL, Bild DE, et al. Surveillance and ascertainment of cardiovascular events: the Cardiovascular Health Study. Ann Epidemiol 1995;5: Mahaffey KW, Granger CB, Tardiff BE, et al. Endpoint adjudication by clinical events committee can impact the statistical outcome of a clinical trial: results from GUSTO-IIb [abstract]. J Am Coll Cardiol 1997;29:410A.

7 248 Mahaffey et al American Heart Journal February Mahaffey KW, Tardiff BE, Granger CB, et al. Disagreement between site investigators and clinical event committees is common and can affect trial results [abstract]. J Am Coll Cardiol 1998;31: 184A-5A. 13. Mahaffey KW, Harrington RA, Akkerhuis M, et al. Disagreements between a central clinical events committee and site investigator assessments of myocardial infarction end-points in an international clinical trial: Review of the PURSUIT study. Curr Control Trials Cardiovasc Med Curr Control Trials Cardiovasc Med 2001;2: Näslund U, Grip L, Fischer-Hansen J, et al. The impact of an endpoint committee in a large multicentre, randomised, placebo-controlled clinical trial. Eur Heart J 1999;20: Moliterno DJ. Patient-specific dosing of IIb/IIIa antagonists during acute coronary syndromes: Rationale and design of the PARAGON-B study. Am Heart J 2000;139: Mahaffey KW, Harrington RA, Akkerhuis M, et al. Systematic adjudication of myocardial infarction endpoints in an international clinical trial. Curr Control Trials Cardiovasc Med 2001;2: IMPACT-II Investigators. Randomised placebo-controlled trial of effect of eptifibatide on complications of percutaneous coronary intervention. Lancet 1997;349: EPIC Investigators. Use of a monoclonal antibody directed against the platelet glycoprotein IIb/IIIa receptor in high-risk coronary angioplasty. N Engl J Med 1994;330: EPILOG Investigators. Platelet glycoprotein IIb/IIIa receptor blockade and low-dose heparin during percutaneous coronary revascularization. N Engl J Med 1997;336: Alpert JS, Thygesen K. 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: Appendix Definition for postenrollment MI: The definition of MI as a study end point took into account whether a patient had an MI at baseline or underwent intervention. MI after enrollment in patients without MI at baseline: Elevation of CK-MB (and 3% of total CK) to 2 upper limit of normal (ULN) or new, significant (0.04 seconds) Q waves in 2 contiguous leads. If CK-MB was unavailable, then total CK >2 ULN was required. MI after enrollment in patients with MI at baseline: If the event in question was <16 hours after the index event, MI was defined by the following (except in patients with percutaneous or surgical intervention during this interval): Severe, recurrent ischemic pain at rest (>30 minutes) AND New or recurrent ST-segment elevation >0.1 mv in 2 contiguous leads. If the event in question was >16 hours after enrollment, MI was defined by the following: Re-elevation of CK-MB to 2 ULN, if prior CK-MB was in the normal range OR CK-MB 2 ULN and >50% above the prior level, if the prior level was above normal OR If CK-MB was unavailable, then total CK >2 ULN and increased by 25% or 200 U/L above the previous value was required OR New, significant (0.04 seconds) Q waves in 2 contiguous leads, discrete from enrollment MI. Periprocedural MI CK-MB (or CK) level 3 ULN and >50% above prior nadir value OR new, significant (0.04 seconds) Q waves in 2 contiguous leads. Perioperative MI CK-MB (or CK) level 5 ULN OR new, significant (0.04 seconds) Q waves in 2 contiguous leads. When enzyme or ECG data were unavailable, an MI was identified when the majority of clinical evidence (patient signs, symptoms, ECG changes, and pathologic findings) so indicated.