Comparison of Long-Term Mortality Across the Spectrum of Acute Coronary Syndromes

Larry A. Allen, MD; Christopher J. O'Donnell, MD, MPH; Carlos A. Camargo Jr., MD, DrPH; Robert P. Giugliano, MD, SM; Donald M. Lloyd-Jones, MD, SM 

Am Heart J.  2006;151(5):1065-1071.  ?2006 Mosby, Inc.
Posted 05/17/2006

Abstract and Introduction

Abstract

Background: Data are sparse regarding comparative long-term mortality across the spectrum of patients presenting with acute coronary syndrome (ACS).
 Methods: We identified all patients hospitalized with suspected myocardial ischemia in an urban academic hospital from 1991 to 1992. We compared presenting characteristics, treatment, and long-term mortality between patients with unstable angina (UA), minor myocardial damage (MMD), definite non–ST-elevation myocardial infarction (NSTEMI), and STEMI.
 Results: Of 760 patients (mean age 68 years, 35% women), 22% had UA, 35% had MMD, 26% had NSTEMI, and 17% had STEMI. During a mean follow-up of 9.5 years, unadjusted mortality was highest in patients with MMD and NSTEMI (mortality for UA 43%, MMD 68%, NSTEMI 62%, STEMI 44%; P < .001). Patients with MMD and NSTEMI were older than patients with STEMI or UA, had more comorbid conditions (diabetes, prior myocardial infarction, congestive heart failure), and were less likely to receive aspirin, unfractionated heparin, or revascularization therapies during the index hospitalization. After multivariable adjustment for all significant covariates, mortality increased sequentially along the spectrum of ACS (hazards ratios for UA 1.0 [referent], MMD 1.12 [95% CI 0.84-1.49], NSTEMI 1.28 [0.95-1.72], and STEMI 1.52 [1.06-2.19]).
 Conclusions: Patients presenting with MMD and definite NSTEMI had a worse unadjusted long-term prognosis up to 10 years after index hospitalization than patients with STEMI. This mortality excess for MMD/NSTEMI was associated with more comorbid conditions and decreased use of basic therapies for ACS. After controlling for baseline differences, STEMI patients had the highest mortality.

Introduction

Despite extensive research examining short-term outcomes in patients with acute coronary syndrome (ACS), the comparative long-term prognoses of ACS subtypes remain relatively unknown. Large databases of myocardial ischemia patients have either neglected to categorize patients using both electrocardiographic findings and serum cardiac marker elevations, or have reported only short duration of follow-up.[1-3] Traditionally, ST-elevation myocardial infarction (STEMI) has been classified as the most severe manifestation of the ACS spectrum.[4,5] However, patients with chest pain and ST depression on electrocardiogram (ECG) have worse 1-year prognosis than their counterparts with ST elevation.[6] Our group previously showed that 1-year mortality after hospitalization for acute myocardial infarction (MI) was higher for patients with unequivocal non–ST-elevation myocardial infarction (NSTEMI), intermediate for patients with minor myocardial damage (MMD, defined by elevated creatine kinase [CK]–MB relative index [MBRI] with normal total CK), and lowest for patients with STEMI.[7] It is unclear whether the poorer prognosis of NSTEMI and MMD, compared with STEMI, is caused by differential pathogenesis, a greater burden of associated comorbidities, or less aggressive care of these patients. We have recently published long-term mortality data for unstable angina (UA)/NSTEMI patients[8,9] and here have extended mortality follow-up out to 10 years in our cohort of STEMI patients to further investigate these issues.

Methods

Measurement of CK and CK-MB

All patients hospitalized with suspected ACS had serial measurements of serum CK and CK-MB every 6 to 8 hours for 24 hours. Total serum CK was measured using an activity assay (Boehringer-Mannheim, Indianapolis, IN; reference range 60-400 IU/L for men, 40-150 IU/L for women, defined according to specifications of the manufacturers). Absolute CK-MB isoenzyme level was measured only in patients with CK level >100 IU/L using a monoclonal antibody–based mass assay (CIBA-Corning Magic-Lite System, Westwood, MA; reference range 0.0-7.5 mg/mL). The MBRI was calculated by dividing the peak total CK-MB level (in nanograms per milliliter) by the associated total CK level (in international units per liter) and multiplying by 100 (reference range 0.0%-3.0%). Troponins were not available for routine clinical practice at our hospital during the index hospitalization.

Study Sample

Methods for patient identification have been described in detail previously.[7,10] Briefly, we screened the emergency department logs and the billing records to identify all adults admitted from the emergency department to the coronary or medical intensive care units or cardiac telemetry unit of a large, urban, teaching hospital between October 1, 1991, and September 30, 1992. Through chart review and using the definition for case identification published in the 1994 Agency for Health Care Policy and Research Unstable Angina Clinical Practice Guideline,[11] we identified 1016 patients with a primary diagnosis of ACS based on their presenting symptoms. To study primary acute coronary syndromes, we excluded 51 patients transferred from other hospitals, 45 patients admitted initially to nonmedical services, 111 patients with precipitants (sepsis, gastrointestinal tract hemorrhage, and severe anemia), 21 patients at high risk for mortality from preexisting illness (cancer, AIDS, and prehospital arrest), 26 patients with nonischemic myocarditis (biopsy-proven with normal coronary arteries), and 2 patients without medical records. The final cohort for this analysis consisted of 760 consecutive nonreferral patients with a primary diagnosis of ACS.

The remaining patients were stratified into 4 mutually exclusive categories. Patients were classified as having STEMI if they had elevated total CK level and MBRI, and =1 mm segment elevation at 80 milliseconds after the J point in =2 contiguous leads on their presenting ECG in the emergency department. The remaining patients were categorized as NSTEMI if they had elevated total CK level and elevated MBRI, as MMD if they had a normal total CK level with elevated MBRI, and as UA if they had a normal MBRI. Although modern definitions of ACS do not formally recognize an MMD subgroup, we felt that distinguishing minor and major cardiac marker release among non–ST-elevation patients would enhance the clinical interpretation of the data presented here.

Baseline Characteristics, Treatments, and Outcomes Data Collection

Baseline data were collected on the study patients by trained physician abstractors through review of paper charts and computer databases as previously described.[7,10] Agreement between chart abstractors was assessed for several key variables (interrater agreement 88%-91%, ? 0.74-0.81, P < .001). Congestive heart failure (CHF) at presentation was defined by the presence of rales or S3 recorded on examination in the emergency department or by chest x-ray consistent with pulmonary edema as read by attending radiologists blinded to the patient's status. Patients were considered to have received appropriate aspirin and intravenous unfractionated heparin therapy if the chart indicated administration within 24 hours of presentation. Patients were considered to have received coronary revascularization if they underwent a percutaneous coronary intervention or a surgical coronary artery bypass procedure at any point during the index hospitalization.

For the present analysis, the primary outcome of interest was long-term mortality after index hospitalization. Vital status data were obtained from hospital records, discharge summaries, laboratory evaluations, and computerized clinic notes. A 1-year follow-up of mortality included contacting some patients by mail or telephone. When vital status remained uncertain through the end of the 10-year follow-up period, searches of the Social Security Death Index and Massachusetts Death Registry were performed; Social Security numbers were available for 97% of patients. This research project was approved by the Subcommittee on Human Studies at our institution.

Statistical Analysis

Initial analyses compared baseline characteristics and outcomes across the UA, MMD, NSTEMI, and STEMI groups using Pearson χ2 and Fisher exact tests for discrete variables, as appropriate, and Kruskal-Wallis tests for continuous variables. For 10-year mortality, Cox proportional hazards models were used to derive hazard ratios. Proportional hazards assumptions were checked and found to be appropriate. For multivariable models, baseline characteristics and inhospital therapies with a univariate 2-tailed P < .15 (across the 4 groups) were considered for inclusion. Stepwise Cox regression models were used to derive multivariable-adjusted hazards ratios for mortality using a P < .05 for retention in the model, after forcing the ACS subtype variable into the model. Because long-term mortality was evaluated, all relevant covariates collected during the index hospitalization were considered for the multivariable-adjusted model, including both presenting characteristics as well as inhospital therapies. For all analyses of the 4 groups, patients with UA served as the reference group. All statistical analyses were performed using STATA version SE 8.0 (STATA Corp, College Station, TX). A 2-tailed P < .05 was considered statistically significant. All hazards ratios are reported with 95% CIs.

Results

Patient Characteristics

Baseline characteristics of the study sample are listed in Table I . The mean age was 68 years, and 35% were women. When stratified by ACS subtype, 22% of the study patients had UA, 35% had MMD, 26% had NSTEMI, and 17% had STEMI. Patients with NSTEMI were much more likely to be women, in part because of different cutoff values for total CK based on sex. Patients with STEMI tended to have less frequent prior history of MI and less baseline use of treatments for coronary artery disease than their counterparts with UA, MMD, and NSTEMI. Patients with STEMI also had relatively lower rates of comorbid conditions (diabetes, history of MI, and CHF), but they had twice the rate of current smoking compared with patients with UA, MMD, and NSTEMI.

Long-Term Mortality Outcomes

Of 760 patients in the study sample, 438 (58%) died during median confirmed follow-up of 9.5 years (interquartile range 2.1-10.1 years). Patients with MMD and NSTEMI had significantly higher unadjusted long-term mortality compared with those with STEMI and UA ( Table II ; unadjusted Kaplan-Meier survival curve for the 4 groups shown in Figure 1). After adjustment for age alone, patients with MMD and NSTEMI remained at higher risk for mortality compared with patients with UA, but the difference was attenuated to nonsignificance between MMD/NSTEMI and STEMI (age-adjusted hazards ratios shown in Table II ; age-adjusted survival curves shown in Figure 2).

Figure 1. 

Unadjusted Kaplan-Meier survival curves up to 10 years from index hospitalization, stratified by type of ACS presentation.

     

Figure 2. 

Age-adjusted survival curves up to 10 years from index hospitalization, stratified by type of ACS presentation.

     

After multivariable adjustment, the hazards for long-term mortality associated with MMD and NSTEMI were substantially attenuated relative to UA, whereas mortality hazards associated with STEMI were relatively increased. Thus, after multivariable adjustment, relative risk for mortality increased in a stepwise fashion along the spectrum of ACS ( Table II ). Covariates selected in the multivariable model included age, prior MI, diabetes, creatinine >1.5 mg/dL, CHF at presentation, appropriate heparin therapy, appropriate aspirin therapy, and coronary revascularization during the index hospitalization; although female sex, prior coronary revascularization, prior ?blocker use, and current smoking were significantly different between groups at baseline in univariate analysis, these additional covariates did not meet criteria for retention in the model and did not significantly affect the results after adjustment for the retained covariates.

Because of the potential confounders introduced by using different cutoffs by sex for normal total CK, we also performed these analyses after redefining UA, MMD, NSTEMI, and STEMI groups using an arbitrary CK cutoff value of 275 IU/L for both men and women. This reclassified 2 men and 58 women between the NSTEMI and MMD groups. The outcome rates and hazard ratios were nearly identical to the prior analyses.

Treatment Differences

In secondary analyses designed to explore the effect of treatment on outcome, patients with MMD and NSTEMI were found to be significantly less likely than patients with STEMI or UA to receive aspirin, unfractionated heparin, or revascularization therapies during the index hospitalization ( Table I ). After adjustment for aspirin, unfractionated heparin, and revascularization therapies alone and in combination, there was significant attenuation of the excess crude mortality associated with MMD or NSTEMI compared with STEMI or UA, but a significant mortality difference persisted between NSTEMI over UA ( Table II ).

Discussion

In our cohort of patients hospitalized with ACS, those with MMD and definite NSTEMI had unadjusted long-term mortality rates approximately 70% higher than patients with STEMI or UA. However, long-term mortality differences between the groups were attenuated after adjustment for comorbidities and treatment during the index hospitalization. After adjusting for these potential confounders, the relative hazards for long-term mortality associated with ACS increased in a stepwise fashion: patients presenting with UA had the lowest risk, patients with MMD and NSTEMI had modestly (but not significantly) higher risk, and patients with STEMI were at highest risk. To our knowledge, these are among the first data comparing long-term mortality outcomes across the spectrum of ACS. These findings extend the results of previous research demonstrating that patients with NSTEMI have unadjusted mortality rates higher than patients with STEMI[6,7] and broaden our understanding of some of the possible explanations for this finding.

Several mechanisms are likely to explain our results. In our cohort of consecutive ACS patients, NSTEMI events (both MMD and definite NSTEMI) occurred more frequently in older patients with prior MI and other comorbidities. Indeed, multivariable adjustment for age and comorbid illnesses in our analysis reversed the hazards ratio trends for mortality between MMD/NSTEMI and STEMI, strongly suggesting that STEMI patients had lower crude mortality because they were younger and healthier at baseline, not because the STEMI event itself was less dangerous.

The persistence of crude mortality differences over many years from the index event is also consistent with the hypothesis that the presenting ACS subtype is a manifestation of fundamental underlying pathophysiological differences between patients. MMD and NSTEMI are more likely than STEMI to occur in the setting of chronic repetitive episodes of short-term arterial occlusion with platelet embolization, preexisting collateral flow in the affected artery, and reinfarction.[12,13] All of these differences reflect a higher baseline level of coronary artery disease in patients with MMD/NSTEMI versus their STEMI counterparts. Subsequent infarcts are known to pose a higher risk than first infarcts.[14]

To explore potential treatment-related effects, we examined the association of 3 proven therapies during the index hospitalization with long-term outcomes. At the time the study sample was collected, patients with MMD and NSTEMI were much less likely to receive basic ACS therapies than their counterparts with STEMI. After adjustment for these treatment differences alone, the mortality differences between ACS subtypes were attenuated and were no longer significant between MMD, NSTEMI, and STEMI. It is possible that the treatment differences themselves contributed to the observed mortality differences over the long term, given the known mortality benefits of these 3 treatments. However, such a lasting effect from therapy delivered at a single point in time seems unlikely. Therefore, it is possible that appropriate care during the index hospitalization predicts continued quality care over the long term.[15]

Alternatively, rates of utilization of ACS treatments may differ not only based on type of ACS but also because of differences in baseline characteristics that sequester with type of ACS. For example, NSTEMI patients were more likely to be women compared with other types of ACS. Numerous studies have demonstrated that women are less likely to receive appropriate therapy for ACS.[16] It should be noted that the higher number of women categorized as NSTEMI in our cohort may in part be caused by the differential cutoff values for normal CK-MB used in this analysis. Regardless, controlling for sex in multivariate analysis and performing our analyses with a single redefined CK cutoff did not significantly affect the results.

More likely, patients with MMD/NSTEMI may have received standard therapies less frequently than patients with STEMI because STEMI can be diagnosed immediately on ECG and has historically been viewed as the most concerning diagnosis in the spectrum of ACS. Furthermore, the presence of greater comorbidities among the MMD/NSTEMI group may have made treating physicians reluctant to pursue therapies with high risk of bleeding or invasive procedures.

Regardless of the cause, patients with MMD/NSTEMI have the potential to benefit greatly from proven medical and invasive therapies because of their high mortality rate. Fortunately, over the past decade since this cohort was assembled, an increasing proportion of ACS patients are viewed as appropriate candidates for aggressive use of antithrombotic therapies, anti-ischemic medications, and revascularization strategies.[4,17]

One possible discrepancy between our findings and prior published data lies in the similar unadjusted mortality of MMD and definite NSTEMI, despite the minimal cardiac marker elevation in the patients with MMD. Multiple studies using large ACS cohorts have shown a strong relationship between absolute cardiac marker elevation and increasing short-term mortality.[18-20] Whereas larger area of infarction is associated with greater risk of acute and subacute mortality, other factors may play a more significant role over the long term. After multivariable adjustment of our data accounting for possible baseline covariates responsible for poor long-term outcomes, the unexpectedly high mortality among MMD patients was markedly attenuated. Again, this supports the conclusion that patients presenting with MMD are more likely to have baseline characteristics that put them at increased risk for long-term mortality relative to patients presenting with other forms of ACS.

An inherent limitation to any study of long-term outcomes is that diagnostic and therapeutic modalities change over the observation period. As a result, we define MMD and NSTEMI using CK-MB rather than through the current standard of troponins.[21-23] Whereas troponins may have somewhat higher sensitivity for detecting small amounts of myocardial necrosis and identify a somewhat different population of MMD than CK-MB,[24] MMD/NSTEMI defined currently by borderline troponin elevations and by the older CK-MB criteria used in our cohort overlap for the majority of patients.[25,26] Definitions and treatment strategies for ACS have also changed considerably over the decade since this cohort of patients underwent their index hospitalization.[4,5] However, the means of identifying patients at presentation and the basic treatments we evaluated in this cohort have remained the foundation of care for patients with ACS.[11] Data regarding the long-term outcomes of patients with minor elevations in troponins will be of interest when available. We would anticipate that such patients will also be at increased risk for long-term mortality, but in the contemporary setting, more widespread application of standard therapies will have resulted in a narrowing of the gap between patients with MMD/NSTEMI and those with STEMI.[3]

Other potential limitations should be recognized. The study was conducted at a single urban academic center, which may limit the generalizability to other settings. Any retrospective chart review is subject to information bias, despite our efforts to use standardized definitions, verify interrater agreement, and obtain mortality data from multiple sources. We cannot exclude residual confounding because of variables that were incompletely controlled for, not collected, or were dependent on the physician's discretion.

Overall, these findings extend previous research indicating that NSTEMI patients have higher crude mortality than STEMI patients and that MMD portends a grave prognosis out of proportion to its relatively limited extent of acute myocardial necrosis. Recent study has focused on the importance of MMD as an indicator for acute risk of subsequent MI and death.[18-20] The significant 10-year mortality we observed places patients with MMD in a high-risk category for long-term mortality as well. Unfortunately, as shown in this older cohort as well as in other more recent analyses, too many patients with ACS in the usual care setting are not receiving basic lifesaving therapies.[27-29] Future efforts to improve outcomes in ACS must include significant efforts to improve adherence with currently available treatments over the short and long terms, particularly in the MMD/NSTEMI population.

Table I. Baseline Characteristics and Treatment Differences


  UA (n = 165) MMD (n = 263) NSTEMI (n = 202) STEMI (n = 130) P*
Mean age (SD), y 66 (12) 69 (12) 71 (12) 62 (12) <.001
Female sex 33% 22% 59% 29% <.001
Prior MI 52% 55% 36% 12% <.001
Prior coronary revascularization 43% 40% 19% 12% <.001
Prior aspirin use 44% 40% 28% 17% <.001
Prior ?blocker use 47% 31% 23% 14% <.001
Current smoking 19% 20% 19% 40% <.001
Diabetes 33% 27% 29% 19% .059
CHF at presentation 26% 31% 38% 23% .013
Serum creatinine >1.5 mg/dL 17% 25% 24% 12% .008
Aspirin therapy within 24 h 78% 68% 59% 89% <.001
Intravenous heparin therapy within 24 h 77% 41% 53% 85% <.001
Revascularization during index hospitalization 35% 13% 16% 45% <.001
Confirmed deaths at 10 y 43% (71) 68% (179) 62% (323) 44% (57) <.001

P values across all 4 groups.

 

Table II. Hazards Ratios for Long-Term Mortality in Patients with MMD, NSTEMI, and STEMI Compared with UA


Model Hazards ratio (95% CI) for long-term mortality
UA MMD NSTEMI STEMI
Unadjusted 1.0 (ref) 1.73 (1.32-2.25) 1.70 (1.28-2.25) 1.02 (0.72-1.44)
Age-adjusted 1.0 (ref) 1.44 (1.10-1.88) 1.34 (1.01-1.78) 1.12 (0.79-1.59)
Multivariable-adjusted* 1.0 (ref) 1.12 (0.84-1.49) 1.28 (0.95-1.72) 1.52 (1.06-2.19)
Aspirin-adjusted only 1.0 (ref) 1.66 (1.27-2.17) 1.60 (1.20-2.13) 1.12 (0.79-1.59)
IV heparin-adjusted only 1.0 (ref) 1.38 (1.05-1.83) 1.55 (1.16-2.06) 1.11 (0.78-1.57)
Revascularization-adjusted only 1.0 (ref) 1.48 (1.13-1.93) 1.59 (1.19-2.11) 1.16 (0.82-1.63)
Adjusted for all 3 therapies 1.0 (ref) 1.27 (0.96-1.68) 1.38 (1.03-1.84) 1.29 (0.91-1.83)

IV, Intravenous.
Covariates: age, prior MI, diabetes, elevated creatinine, CHF at presentation, appropriate heparin therapy, appropriate aspirin therapy, and revascularization therapy.

 



References

  1. Jacobs DR, Kroenke C, Crow R, et al.. PREDICT: a simple risk score for clinical severity and long-term prognosis after hospitalization for acute myocardial infarction and unstable angina: the Minnesota Heart Study. Circulation. 1999;100:599–607.
  2. Gillum RF. Trends in acute myocardial infarction and coronary heart disease death in the United States. J Am Coll Cardiol. 1994;23:1273–1277.
  3. Engstrom G, Goransson M, Hansen O, et al.. Trends in long-term survival after myocardial infarction. J Intern Med. 2000;248:425–434.
  4. Braunwald E, Antman EM, Beasley JW, et al.. ACC/AHA guideline for the management of patients with unstable angina and non–ST-segment elevation myocardial infarction: 2002 update: 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). J Am Coll Cardiol. 2002;40:1366–1374..
  5. Antman EM, Anbe DT, Armstrong PW, et al.. ACC/AHA guidelines for the management of patients with ST-elevation myocardial infarction—executive summary. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (writing committee to revise the 1999 guidelines for the management of patients with acute myocardial infarction). J Am Coll Cardiol. 2004;44:671–719.
  6. Nyman I, Areskog M, Areskog N, et al.. Very early risk stratification by electrocardiogram at rest in men with suspected unstable coronary heart disease. The RISC study group. J Intern Med. 1993;234:293–301.
  7. Lloyd-Jones DM, Camargo CA, Giugliano RP, et al.. Characteristics and prognosis of patients with suspected acute myocardial infarction and elevated MB relative index but normal total creatinine kinase. Am J Cardiol. 1999;84:957–962.
  8. Allen LA, O'Donnell CJ, Giugliano RP, et al.. Care concordant with guidelines predicts decreased long-term mortality in patients with unstable angina pectoris and non–ST-elevation myocardial infarction. Am J Cardiol. 2004;93:1218–1222.
  9. Lloyd-Jones DM, Camargo CA, Allen LA, et al.. Predictors of long-term mortality after hospitalization for primary unstable angina pectoris and non–ST-elevation myocardial infarction. Am J Cardiol. 2003;92:1155–1159.
  10. Giugliano RP, Camargo CA, Lloyd-Jones DM, et al.. Elderly patients receive less aggressive medical and invasive management of unstable angina: potential impact of practice guidelines. Arch Intern Med. 1998;158:1113–1120.
  11. Braunwald E, Mark DB, Jones RH, et al.. Unstable angina: diagnosis and management. Clinical practice guideline number 10. Bethesda (Md): Agency for Health Care Policy and Research, Public Health Service, National Heart, Lung, and Blood Institute; 1994;.
  12. Davies MJ. Coronary disease: the pathophysiology of acute coronary syndromes. Heart. 2000;83:361–366.
  13. Phibbs B, Marcus F, Marriott HJC, et al.. Q-wave versus non–Q-wave myocardial infarction: a meaningless distinction. J Am Coll Cardiol. 1999;33:576–582.
  14. Mueller H, Forma S, Menegus M, et al.. Prognostic significance of nonfatal reinfarction during 3-year follow-up results of TIMI phase II clinical trial. J Am Coll Cardiol. 1995;26:900–907.
  15. Bhatt DL, Roe MT, Peterson ED, et al.. Utilization of early invasive management strategies for high-risk patients with non–ST-elevation acute coronary syndromes: results from the CRUSADE quality improvement initiative. JAMA. 2004;292:2096–2104.
  16. Gan CG, Beaver SK, Houck PM, et al.. Treatment of acute myocardial infarction and 30-day mortality among women and men. N Engl J Med. 2000;343:8–15.
  17. Cannon CP, Weintraub WS, Demopoulos LA, et al.. Comparison of early invasive and conservative strategies in patients with unstable coronary syndromes treated with the glycoprotein IIb/IIIa inhibitor tirofiban (TACTICS-TIMI 18). N Engl J Med. 2001;344:1879–1887.
  18. Alexander JH, Sparapani RA, Mahaffey K, et al.for the PURSUIT Steering Committee. Association between minor elevations of creatine kinase–MB level and mortality in patients with acute coronary syndromes without ST-segment elevation. JAMA. 2000;283:347–353.
  19. Antman EM, Tanasijevic MJ, Thompson B, et al.. Cardiac-specific troponin I levels to predict the risk of mortality in patients with acute coronary syndromes. N Engl J Med. 1996;335:1342–1349.
  20. Morrow DA, Cannon CP, Rifai N, et al.for the TACTICS-TIMI 18 Investigators. Ability of minor elevations of troponins I and T to predict benefit from an early invasive strategy in patients with unstable angina and non–ST elevation myocardial infarction: results from a randomized trial. JAMA. 2001;286:2405–2412.
  21. Antman E, Bassand JP, Klein W, et al.. 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;36:959–969.
  22. World Health Organization. Manual of the international statistical classification of diseases, injuries, and causes of death, based on the recommendations of the ninth revision conference, 1975. Geneva (Switzerland): World Health Organization; 1977;.
  23. Newby LK, Storrow AB, Gibler WB, et al.. Bedside multimarker testing for risk stratification in chest pain units: the Chest Pain Evaluation by Creatine Kinase–MB, Myoglobin, and Troponin I (CHECKMATE) study. Circulation. 2001;103:1832–1837.
  24. Roger VL, Killian JM, Weston SA, et al.. The new criteria for the diagnosis of myocardial infarction: prospective evaluation in the community. Abstract. AHA Scientific Sessions, November 10, 2004. Accessed at http://scientificsessions.americanheart.org on March 2, 2005.
  25. Adams JE, Schechtman KB, Landt Y, et al.. Comparable detection of acute myocardial infarction by creatine kinase MB isoenzyme and cardiac troponin I. Clin Chem. 1994;40:1291–1295.
  26. Harris BM, Nageh T, Marsden JT, et al.. Comparison of cardiac troponin T and I and CK-MB for the detection of minor myocardial damage during interventional cardiac procedures. Ann Biochem. 2000;37:764–769.
  27. Jencks SF, Huff ED, Cuerdon T. Change in the quality of care delivered to Medicare beneficiaries, 1998-1999 to 2000-2001. JAMA. 2003;289:305–312.
  28. Marciniak TA, Ellerbeck EF, Radford MJ, et al.. Improving the quality of care for Medicare patients with acute myocardial infarction: results from the cooperative cardiovascular project. JAMA. 1998;279:1351–1357.
  29. Krumholz HM, Chen J, Rathore SS, et al.. Regional variation in the treatment and outcomes of myocardial infarction: investigating New England's advantage. Am Heart J. 2003;146:242–249.
Reprint Address

Donald M. Lloyd-Jones, MD, SM, Division of Cardiology, Northwestern University, 680 N Lake Shore Drive, Suite 1102, Chicago, IL 60611. E-mail: dlj@northwestern.edu


Larry A. Allen, MD,a Christopher J. O'Donnell, MD, MPH,b,e Carlos A. Camargo Jr., MD, DrPH,c Robert P. Giugliano, MD, SM,d Donald M. Lloyd-Jones, MD, SM,f

aDivision of Cardiology, Department of Medicine, Duke University Medical Center, Durham, NC; bCardiology Division, Department of Medicine, Massachusetts General Hospital and Harvard Medical School, Boston, MA; cDepartment of Emergency Medicine, Massachusetts General Hospital and Harvard Medical School, Boston, MA; dCardiovascular Division, Department of Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, MA; eNational Heart, Lung, and Blood Institute's Framingham Heart Study, Framingham, MA; fDepartment of Preventive Medicine and Division of Cardiology, Feinberg School of Medicine, Northwestern University, Chicago, IL

Disclosure: Dr Lloyd-Jones is supported by grant K23 HL04253 from the National Heart, Lung, and Blood Institute.