Original Contributions

Defining the outcomes of risk stratification studies of ED patients with chest pain: the marginal value of adding revascularization to the Composite end point

Esther H. Chen MD*, Frank Sites RN, MHA, Frances S. Shofer PhD, Judd E. Hollander MD

Department of Emergency Medicine, Hospital of the University of Pennsylvania, Philadelphia, PA 19104-4283, USA

Abstract

Objectives: Cardiovascular risk stratification studies use various end points, sometimes including revascularization. We assessed whether adding revascularization to a strictly defined composite end point of death, acute myocardial infarction (AMI), and Unstable angina impacts the likelihood of patients attaining the composite end point.

Methods: We conducted a secondary analysis of a prospectively collected data set of emergency department patients who received an electrocardiogram for chest pain. Patients were followed daily; discharged patients had 30-day telephone follow-up. The main outcome was a 30-day composite end point of death, AMI, and UA compared with death, AMI, UA, and revascularization.

Results: There were 4492 patients enrolled (mean age, 52 F 16 years; men, 41%; African American, 68%). One hundred seventy patients were revascularized (158 had AMI or UA). Overall, the incidence of death/AMI/UA was 20.1% (95% confidence interval, 18.9%-21.2%). With revascularization included, the incidence of the composite end point was 20.3% (95% confidence interval, 19.1%-21.5%). Conclusion: When both AMI and UA are strictly defined, there appears to be a limited role for adding revascularization to a composite end point of death, AMI, and UA because most revascularized patients have a diagnosis of AMI or UA.

D 2005

Introduction

Patients with acute chest pain represent about 5.3% of all emergency department (ED) visits [1] and challenge the emergency physician to identify, treat, and prevent adverse events in high-risk patients, as well as to reduce unnecessary hospital admissions in low-risk patients. In

T Corresponding author. Tel.: +1 215 349 8506; fax: +1 215 662 3953.

E-mail address: [email protected] (E.H. Chen).

response to emerging clinical trials and epidemiological studies, it became necessary to standardize outcome measures. Management guidelines for acute coronary syndromes (ACSs) were recently updated in attempts to incorporate a wide range of studies into a comprehensive Evidence-based approach [2,3]. However, interpretation of results is difficult because inclusion criteria, historical parameters, and outcome measures vary among studies [4]. Subsequently, the Emergency Medicine Cardiac Re- search and Education Group initiated an effort to standard- ize operational definitions and reporting criteria to facilitate

0735-6757/$ - see front matter D 2005 doi:10.1016/j.ajem.2005.04.006

cross-study comparisons. These proposed criteria have recently been endorsed by the Society of Academic Emer- gency Medicine, American College of Emergency Physi- cians, American Heart Association, and American College of Cardiology [5].

After these standardization efforts, it is useful to evaluate the importance of various Clinical end points used in ACS studies. Risk stratification studies of ED patients with chest pain have used death, acute myocardial infarction (AMI), and unstable angina (UA) as clinical outcome measures, whereas others also include revascularization in the com- posite end point. In the Standardized reporting guidelines for studies evaluating risk stratification of ED patients with potential ACSs [5], revascularization was not considered a mandatory criterion to be reported. Therefore, we ques- tioned whether the absence of this criterion would impact the results of these studies. We hypothesized that the addition of revascularization to a composite end point of death, AMI, and UA would not alter the likelihood of achieving a negative composite end point.

Materials and methods

Study design

This was a secondary analysis of a prospectively collected data set of ED patients with chest pain. We assessed whether adding revascularization to a strictly defined composite end point of death, AMI, and UA impacts the likelihood of patients attaining the composite end point. The study was approved by the University of Pennsylvania Committee on Research Involving Human Subjects.

Study setting and population

Patients were enrolled at an urban tertiary care university hospital ED with an annual census of approximately 51000 visits during the study period, from July 1999 to March 2002. Patients were included if they were older than 24 years with chest pain prompting an electrocardiogram (ECG) and younger than 24 years only if they used cocaine. Broad inclusion criteria were intentionally chosen to ensure generalizability of the data.

Study protocol and measurements

Trained research assistants enrolled ED patients 16 hours per day, 7 days per week, which captures 85% to 95% of eligible patients. Information collected for each patient included demographics, historical description of chest pain characteristics and associated symptoms, laboratory data, and ECG data. Hospitalized patients were followed daily for complications and interventions. Clinical information was obtained from the treating physician. Determination of final diagnosis and cardiac complications was made during daily communication between the investigators and the health- care team. Postdischarge medical record review was not

used. A 30-day follow-up for all study subjects was obtained by a standardized telephone interview.

• • 1. Cardiac biomarker assays

Venous blood samples were collected on presentation to the ED in phlebotomy tubes containing no anticoagulant or preservative. Cardiac troponin I and creatine kinase-MB (CK-MB) were measured by an enzyme-linked immunoas- say using an Abbott AxSYM automated analyzer (Abbott Laboratories, Mountain View, Calif).

• • 1. Main outcome

The main outcome measure was the 30-day composite end point of death, AMI, and UA compared with the composite end point of death, AMI, UA, and revascularization.

Death was defined as all-cause mortality. AMI was defined in accordance with the European Society of Cardiology/American College of Cardiology 2000 consen- sus definition [6]. A diagnosis of AMI was made if the patient had an elevation of cardiac troponin I of 2 ng/mL or greater or CK-MB enzyme of 10 ng/mL or greater. UA was defined using standardized criteria [5,7]. UA was considered to occur if there was a documented reversible ischemia on stress test, coronary artery occlusion of 70% or higher in at least 1 vessel as seen during cardiac catheterization, or elevations of cardiac enzymes greater than laboratory normal but less than levels necessary for diagnosis of AMI (troponin I, z0.4 ng/mL but b2 ng/mL; CK-MB, z5ng/mL but b10 ng/mL) [7]. We did not use the Canadian Cardiovascular Society or Agency for Health Care Policy and Research Clinical Practice Guideline descriptions of UA because they assume a diagnosis of ischemic chest pain and do not apply to unselected ED patients with chest pain. Revascularization was defined as percutaneous coronary intervention (PCI) and coronary artery bypass graft (CABG).

Data analysis

For patients with multiple visits, presentation character- istics and outcomes were counted separately for each visit, but age, race, and sex were counted only once. Outcome data are presented with percentage of frequency of occurrence, with 95% confidence intervals (CIs) provided for main outcomes. Data were analyzed using SAS statistical software (Version 8.2; SAS Institute, Cary, NC).

Results

There were 3819 patients who presented to the ED 4492 times during the study. Patients had a mean age of

51.8 F 17.4 years and were more likely to be women (59%) and African American (68%). Table 1 shows their demo- graphic and historical characteristics. Of the total cohort, 1674 patients were discharged to home (37%); the remainder was admitted to the hospital. There were 2147 patients admitted to telemetry (48%); 496 admitted to the intensive care unit (11%); 135 admitted to unmonitored

outcome did not alter the overall likelihood of reaching the Composite outcome.

Table 1 Characteristics of the patient population

Characteristic No. of study patients (n = 4492 visits)

Sex^a Male Female

cardiac risk factors

1550 (41)

2269 (59)

Values are presented as mean F SD or n (%). CAD indicates coronary artery disease.

^a Age, sex, and ethnicity are only counted once per patient (n = 3819) regardless of number of visits.

Age (y)	51.8 F 15.9
Ethnicitya
Asian	69 (2)
Black	2613 (68)
Hispanic	43 (1)
Other/unknown	31 (1)
White	1063 (28)

Hypertension	2195 (49)
Family history of premature CAD	903 (20)
Diabetes mellitus	806 (18)
Tobacco use	1685 (38)
Hypercholesterolemia/hyperlipidemia	789 (18)
cocaine use	104 (2)
Prior myocardial infarction	497 (12)
known CAD	878 (20)
Overall ECG impression Normal	1912 (43)
Early repolarization only	171 (4)
Nonspecific/nondiagnostic changes	2085 (50)
Ischemia	224 (5)
AMI	100 (2)

floor beds (3%); 14 admitted directly to the catheterization laboratory without another inpatient bed assignment (b1%);

13 patients signed out against medical advice, and 10 patients were transferred to another hospital. Two patients died before admission and 1 patient left before a complete evaluation could be performed. Thirty-day follow-up information was obtained for 98% of the study patients.

During the hospitalization, 319 patients sustained an AMI (7%) and 545 patients were diagnosed with UA (12%). In addition, 29 patients died, 11 developed a late myocardial infarction, 128 had PCI, and 26 received CABG (4 patients received PCI before CABG). At 30-day follow-up, 20 patients sustained an AMI, 17 patients had PCI, 11 patients received CABG, and 54 patients died. Eight of the revascularized patients also had PCI during the index hospitalization.

With respect to the main outcome, 170 patients had revascularization performed. Of these, 158 had a diagnosis of AMI or UA during initial hospitalization. Overall, the incidence of the triple composite end point was 20.1% (95% CI, 18.9%-21.2%). When revascularization was added to this composite outcome, the incidence of the composite end point was 20.3% (95% CI, 19.1%-21.5%). Thus, the addition of revascularization to the triple composite

Discussion

EPs are unable to accurately identify patients at Very low risk for ACS. More than 2% of patients with ACS are sent home from the ED, with a higher 30-day mortality compared with hospitalized patients [8]. These patients are more likely to present with Atypical features and have normal or nondiagnostic ECGs. To reduce the number of missed diagnoses and standardize emergency care of patients with chest pain, practice guidelines that incorporate risk stratification tools and current therapies were developed [2,3]. Unfortunately, these expert recommendations have not been uniformly adhered to, and these risk stratification tools have not been widely accepted.

Several solutions have been proposed to solve the problem of noncompliance with expert guidelines and risk stratification algorithms. For example, the CRUSADE initiative takes a multidisciplinary approach to improving the care of patients with ACS by encouraging adherence to guidelines, implementing ED-focused educational interven- tions, and providing a national registry of high-risk patients with chest pain, a tool for continuous quality improvement [9]. In addition, Emergency Medicine Cardiac Research and Education Group, an international collaboration of emer- gency health-care professionals, has taken a multifaceted approach to improving the diagnosis and treatment of ACS by involving EPs in establishing clinical pathways, encour- aging collaborations with cardiologists in clinical trials, and standardizing reporting criteria in cardiovascular research. Standardizing reporting criteria of operational and outcome definitions enables EPs to perform more rigorous scientific research and to compare clinical outcomes of patients in trials that use various risk stratification tools [5].

This study specifically addressed whether adding revas- cularization to a strictly defined triple composite end point of death, AMI, and UA will alter the incidence of patients who reach a composite outcome. We found that the likelihood of attaining the outcome was the same with or without the use of revascularization in the composite index. The most obvious explanation for this finding is that most patients who received revascularization had a diagnosis of AMI or UA during their hospital visit. In fact, this was true for 158 (93%) of the 170 patients who received revascularization.

Limitations

Our study population was predominantly women and African American. Therefore, our results may not generalize to other patient populations. Furthermore, because this was a single institution study, differences in revascularization rates (local, regional, sex, and ethnic) could not be assessed and may further limit the generalizability of this study.

Although we used standard definitions for the adverse outcomes, we were unable to standardize the interventional strategies for our patients and relied upon the clinical judgment of our cardiologists.

We attempted to reduce the limitations that are pervasive in studies of ED patients with chest pain. Trained research assistants enrolled patients prospectively into the study to reduce selection bias. standardized definitions of adverse outcomes were used and strictly enforced. Inhospital tracking was performed daily rather than relying on medical record review, which reduced misclassification bias.

Conclusions

There appears to be a limited role for adding revascu- larization to a triple composite end point of death, AMI, and UA when these items are strictly defined because most revascularized patients have a diagnosis of AMI or UA.

References

1. McCraig LF, Burt CW. National hospital ambulatory medical care survey: 2001 emergency department summary Advanced Data from Vital and Health Statistics. No. 335, 1.7MG. Hyattsville (Md)7 National Center for Health Statistics; 2003.
2. Braunwald E, Antman E, Beasley J, et al. ACC/AHA guideline update for the management of patients with unstable angina and non-ST- segment-elevation myocardial infarction: 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(7):1366 - 74.
3. Ryan T, Antman E, Brooks N, et al. 1999 update. ACC/AHA guidelines for the management of patients with acute myocardial infarction: executive summary and recommendations. Circulation 1999;100: 1016 - 30.
4. Glavan B, Shewarkramani S, Hollander J. Incomplete data reporting in studies of emergency department patients with potential acute coronary syndromes using troponins. Acad Emerg Med 2003;10(9):943 - 8.
5. Hollander JE, Blomkalns AL, Brogan Jr GX, et al. Standardized reporting guidelines for studies evaluating risk stratification of ED patients with potential acute coronary syndromes. Acad Emerg Med 2004;11:1331 - 40.
6. 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(3):959 - 69.
7. Hollander JE, Mayorga O, Sites FD, Shofer FS. Defining the outcomes of ED chest pain patient risk stratification studies: use of a standardized definition for unstable angina [Abstract]. Acad Emerg Med 2002; 9(5):375.
8. Pope J, Aufderheide T, Ruthazer R, et al. Missed diagnoses of Acute cardiac ischemia in the emergency department. N Engl J Med 2000;342:1163 - 70.
9. Hoekstra J, Pollack C, Roe M, et al. Improving the care of patients with non-ST-elevation acute coronary syndromes in the emergency depart- ment: the CRUSADE initiative. Acad Emerg Med 2002;9(11):1146 - 55.