Original Contribution

Impact of a negative prior stress test on emergency physician Disposition decision in ED patients with chest pain syndromes

Rebecca H. Nerenberg BS, Frances S. Shofer PhD, Jennifer L. Robey RN, BSN, Aaron M. Brown BS, Judd E. Hollander MD*

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

Received 24 April 2006; revised 26 May 2006; accepted 28 May 2006

Abstract

Objective: Many emergency department (ED) patients with potential acute coronary syndromes have prior visits and prior cardiac testing; however, the effect of knowledge of prior testing on the emergency physician disposition decision making is not known. We studied the impact of prior noninvasive testing (ie, Stress testing) for myocardial ischemia on disposition decision making in ED patients with potential ACS.

Methods: We performed a prospective cohort study of ED patients with chest pain who received an

electrocardiogram for potential ACS. Data included demographics, medical history, stress test history, and TIMI risk score. Patients were followed in-house; 30-day telephone interviews were performed for follow-up. Main outcomes were ED disposition (admit/discharge) and a composite of 30-day death, acute myocardial infarction, and revascularization stratified on the basis of prior stress testing known at the time of presentation. Standard statistical techniques were used with 95% confidence intervals (CI). Results: There were 1853 patients enrolled and 97% had follow-up. Patients had a mean age of 53 F 14 years; 60% were women, 67% were black. There were 1491 (79%) patients without a prior stress test, 291 (16%) had a normal prior stress test result, and 89 (5%) had an abnormal prior stress test result. Admission rates were 92% (95% CI, 87%-98%) for patients with a prior abnormal stress test, 73% (95% CI, 67%-78%) for patients with a normal prior stress test, and 70% (95% CI, 67%-72%) for patients without a prior stress test. Adverse outcomes were the highest among patients with prior abnormal stress test but did not differ significantly between patients with no prior stress test and patients with prior normal stress test (10.1% [95% CI, 3.6-16.7%] vs 5.2% [95% CI, 4.1-6.4%] vs 4.8%

[95% CI, 2.4-7.3%]).

Conclusion: Patients without prior stress tests and patients with prior normal stress tests were admitted for potential ACS at the same rate and had the same 30-day cardiovascular event rates. This suggests that prior stress testing does not affect subsequent Disposition decisions. Perhaps cardiac catheterization or computed tomography coronary angiograms would have more of an impact on subsequent visits, making them potentially more cost-effective in the low-risk patient.

D 2007

Presented at the SAEM 2006 annual meeting, San Francisco, CA, May 2006.

* Corresponding author. Tel.: +1 267 250 7167; fax: +1 215 662 3953.

0735-6757/$ - see front matter D 2007 doi:10.1016/j.ajem.2006.05.027

Introduction

Each year, approximately 5.8 million people present to the ED with complaints of chest pain, representing 5% of all ED visits [1]. The decision to admit patients for potential acute coronary syndrome (ACS) is not always clear, resulting in hundreds of thousands of unnecessary admis- sions, whereas approximately 2% of patient with acute myocardial infarction (AMI) are inadvertently discharged from the ED [2]. Many strategies have been implemented over the years, both to decrease unnecessary admissions and to avoid missing patients with ACS [3-10].

One Diagnostic modality commonly used to evaluate patients with potential ACS is an exercise or pharmacologic stress test. The result of the stress test helps brule inQ or brule outQ ACS during any given admission. But what impact do these stress tests have on the patient’s next ED visit? Are they used by physicians to assist with subsequent disposi- tion decision making? Does a previous negative stress test have any prognostic value for a future ED visit?

Chan et al [11] evaluated patients admitted to monitored telemetry beds who received inpatient vs outpatient vs no stress testing and found no difference in 30-day outcome measures. Tatum et al [7] looked at the utility of ED myocardial perfusion imaging in low-risk patients for ACS, and Lai et al [12] studied the necessity for inpatient stress testing for low-risk patients. Smith et al examined the incidence of AMI in patients with a documented negative stress test within the previous 3 years and found that 4.8% of them were diagnosed with AMI [13]. This implies that a recent negative stress test does not conclusively rule out AMI when a patient has a new episode of symptoms.

Given that stress testing is often used in the evaluation of patients with chest pain, it would be expected that prior stress test results may impact ED physician’s decision making during subsequent visits. There are no data, however, about what impact, if any, physician knowledge of a previous stress test has on the disposition decision making and therefore on 30-day outcome. The goal of this study was to determine whether ED physician knowledge of a previous stress test impacted disposition decision making for patients with potential ACS.

Methods

Study design

We performed a prospective cohort study to determine what impact, if any, knowledge of a previous stress test has upon emergency physician disposition decision making during a subsequent ED visit for potential ACS.

Setting

This study was conducted in the ED of a university- affiliated, urban tertiary care center with an ED census of more than 55000 visits per year.

Participants

Consecutive patients aged 30 years or older who presented to the ED between July 13, 2003, and May 31, 2005, with chest pain who received an electrocardiogram were enrolled. Patients who were younger than 30 years, admitted cocaine use, and those with ST segment elevation myocardial infarction (STEMI) were excluded, as these cohorts would be unlikely to have heavy reliance on a prior stress test.

During the study period, trained research assistants were present in the ED from 8 am to midnight 7 days per week to identify and enroll appropriate patients. Written consent for study participation was obtained. Patients were treated by board-certified emergency physicians working in conjunc- tion with house staff. Patient treatment was at the discretion of the treating physician and was not influenced by study enrollment. Data collection was approved by the institu- tional review board.

Measurements

Patients received a structured history and physical examination as well as an electrocardiogram, with all information collected prospectively in the ED. The struc- tured history included questions about the onset of chest pain, the number of previous episodes of chest pain, and the character of the pain. Patients were queried about Family history of MI before 55 years of age, medical history of hypercholesterolemia, hypertension, coronary artery disease (CAD), diabetes mellitus, angina, and congestive heart failure as well as prior stress testing, Coronary artery bypass grafting (CABG), and cardiac catheterization. A TIMI risk score was calculated for each patient by the treating physician. Patient care including disposition was determined by the treating physician and recorded as part of the data set. Physicians were blinded to study hypothesis. Data were collected in accordance with the Standardized reporting guidelines [14]. Admitted patients were followed up daily by one of the investigators. Thirty-day cardiovascular outcomes were obtained via follow-up telephone calls to patients or their contacts. The national death index was searched for patients that could not be contacted for 30-day follow-up.

Noninvasive (stress) testing

Patients who reported prior stress tests from our institution generally received an exercise test with either echocardiography or sestamibi, or a dobutamine or persan- tine stress test with sestamibi. Exercise treadmill testing without imaging is uncommon at our institution.

During the hospitalization, when patients received stress testing, patients were injected with 10 mCi technetium Tc 99m sestamibi before getting resting tomographic images of the heart using single-photon tomographic techniques. Approximately 2 hours later, they underwent the stress portion of the test. exercise stress tests were performed on a treadmill test using a Bruce or modified Bruce protocol.

Alternatively, patients received an infusion of 0.142 mg/kg per minute of dipyramidole over 4 minutes, or 5 lg/kg per minute of dobutamine increasing to 10, 20, and	Table 2 Disposition by prior stress test Disposition Prior stress test
30 lg/kg per minute over 3-minute intervals. Approximate- ly 30 minutes after the exercise or infusions, tomographic	Admitted (95% CI) (%)	Abnormal 92 (87-98)	None 67 (67-72)	Normal 72 (67-78)
imaging of the heart was again obtained. A board-certified/	Discharged (%)	8	30	27
eligible cardiologist interpreted the test. For purposes of this	Left against	0	2	2
study, the stress test was considered positive if there were signs of reversible ischemia. Nondiagnostic tests were those	medical advice (%)

that did not achieve a heart rate 80% of predicted maximum and those that had equivocal images (eg, an inability to distinguish breast attenuation from inferior ischemia).

Outcome measures

The primary outcome was the emergency physician disposition decision. The secondary outcome was a com-

Table 1

Patient characteristics

Prior stress test Abnormal None

P

Normal

Associated symptoms

Chest pain characteristics (median, min)

Values are presented as number (%) unless otherwise indicated.

posite of 30-day cardiovascular events including death, nonfatal AMI, percutaneous coronary intervention, or CABG. AMI was defined as a troponin I level greater than 2.0 or creatine kinase-MB (CK-MB) level greater than 10, in accordance with institutional norms. Self (or proxy)- reporting of AMI, percutaneous coronary intervention, CABG, or death was recorded at 30-day follow-up.

Acute coronary syndrome included AMI and unstable angina. Unstable angina was objectively defined in accor- dance with the standardized reporting guidelines as those patients with consistent symptoms and proven underlying coronary disease including a positive exercise stress test (N1.5 mm ST-segment depression), reversible Ischemic changes on stress echocardiogram or sestamibi scan, 70% or greater blockage of at least one coronary artery on angiography, or elevation of cardiac TnI or CK-MB level above normal but less than the limits set for AMI (0.3 ng/mL V cardiac TnI b 2 ng/mL; or 5 ng/mL V CK-MB b 10 ng/mL) [14].

Patient characteristics
Age (mean, y)	60 F 10	52 F 14	54 F 13	b.05
Sex
Female	39 (43)	900 (62)	164 (56)	b.05
Male	51 (57)	552 (38)	131 (44)
Race
African American	58 (64)	974 (67)	192 (65)	.57
White	26 (29)	425 (29)	93 (32)
Other	6 (7)	52 (4)	10 (3)
cardiac risk factors
Hypertension	76 (85)	740 (52)	178 (61)	b.05
Diabetes mellitus	32 (36)	285 (20)	61 (21)	b.05
Hypercholesterolemia	48 (54)	307 (22)	104 (36)	b.05
Family history of	19 (21)	208 (15)	43 (15)	.23
MI at age b55 y Tobacco use	48 (54)	513 (36)	109 (37)	b.05
Undiagnosed chest pain	10 (11)	127 (9)	60 (21)	b.05
Prior cardiac events Congestive heart failure	21 (24)	149 (10)	36 (12)	b.05
CAD	68 (76)	224 (16)	66 (23)	b.05
Angina	50 (56)	171 (12)	51 (17)	b.05
MI	38 (43)	162 (11)	44 (15)	b.05

Primary data analysis

Data are presented as means F SD for continuous variables and frequency and percent for categorical varia- bles. To determine whether patients differed by stress test category (abnormal, normal, or none) and baseline characteristics and outcomes (AMI, ACS, or 30-day events), analysis of variance and the v² test was used for continuous and categorical data, respectively. Assuming a 70% admis- sion rate in patients without a prior stress test, this study had 90% power to detect a 10% absolute decrease in the

Table 3 Admission by location and prior stress test

Location Prior stress test

Abnormal None Normal

Values are presented as number (%).

ICU, indicates intensive care unit; CCU cardiac care unit; CICU, cardiac intermediate care unit; MICU, medical intensive care unit; SICU, surgical intensive care unit.

Shortness of breath	52 (58)	763 (53)	144 (49)	.28
Diaphoresis	24 (27)	281 (19)	61 (21)	.23
Nausea	24 (27)	290 (20)	65 (22)	.25
Vomiting	8 (9)	89 (6)	15 (5)	.42
Lightheadedness	21 (23)	252 (17)	44 (15)	.18
Syncope	0 (0)	36 (2)	2 (1)	.05
Palpitations	11 (12)	181 (12)	37 (13)	1.0
Chest pain at rest	78 (87)	1196 (82)	247 (84)	1.44
Multiple episodes	49 (54)	690 (48)	148 (50)	.33
of chest pain

Cardiac catheterization	2 (2)	4 (.28)	0 (0)
laboratory
ICU (includes CCU,	21 (23.6)	107 (7.5)	20 (6.87)
CICU, MICU, SICU)

Telemetry bed	54 (60.7)	771 (54.3)	73 (59.5)
Nonmonitored bed	5 (5.6)	70 (4.93)	17 (5.84)
Discharged	7 (7.87)	431 (30.3)	79 (27.2)
Other	0 (0)	38 (2.67)	2 (0.69)

Chest pain onset	180	240	240	.19
Chest pain duration	120	120	180	.70

Table 4 Discharge diagnoses by prior stress test
Final diagnosis	Prior stress test
	Abnormal, No. (%) 95% CI	None, No. (%) 95% CI	Normal, No. (%) 95% CI
ACS	22 (25.0) 13.0-34.0	130 (8.7) 7.3-10.1	31 (10.8) 7.2-14.4
AMI	9 (10.2) 3.9-16.5	62 (4.2) 3.2-5.2	12 (4.2) 1.9-6.5
CP NOS	39 (44.3) 33.9-54.7	988 (66.4) 64.0-68.8	190 (66.0) 60.5-71.5
Nonischemic	18 (20.5) 12.1-29.3	204 (13.7) 12.0-15.4	53 (18.4) 13.9-22.9
Definitive other diagnosis	0 (0) 0-3.4	17 (1.1) 0.6-1.6	2 (0.7) 0.1-4.0
CP NOS, chest pain not otherwise specified.

admission rate for patients with a prior negative stress test (PASS 2002, Jerry Hintze, Number Cruncher Statistical System, Kaysville, UT). All comparisons were performed using SAS statistical software (Version 9.1, SAS Institute, Cary, NC).

Results

During the 21 months of the study, 1853 eligible participants were enrolled. The average patient age was 53 F 14 years; 60% were women, and 67% were Black. Patient characteristics are presented in Table 1.

Of the 1853 enrolled patients, 295 (16%) had a documented normal stress test before presentation,

90 (4.9%) had a prior documented abnormal stress test. One thousand four hundred ninety-one (79%) patients had no prior stress test. Disposition by prior stress test, admission location by prior stress test, and discharge diagnoses are presented in Tables 2 - 4.

Thirty-day follow-up was available on 1801 (97%) patients. Adverse outcomes (including AMI and death) were significantly higher for patients with a prior abnormal stress test compared with either prior normal or no prior stress test ( P b .0001; Table 5). However, there was no difference in adverse outcome rate between patients with a prior normal stress test and patients without a prior stress test.

Discussion

Most low-risk patients with chest pain are admitted to the hospital or observed in chest pain units to rule out an ACS [2,4,5,7,9,11,15]. Typically, this cohort of patients receives predischarge provocative testing to risk-stratify patients in terms of prognosis and to assist further management. Previous studies have evaluated the usefulness and cost-effectiveness of stress testing in the ED, the chest pain observation unit, as an inpatient and as an outpatient, and the impact stress testing has on repeat ED visits [12,13,15-19].

Several studies have shown that, in actual clinical practice, the minority of patients (21%-22%) admitted to a non-intensive care telemetry bed to rule out an ACS

received a stress test before discharge and that testing is relatively low yield [11,18]. Amsterdam et al found that only 13% of 1000 low-risk patients receiving immediate exercise testing in the ED had a positive exercise test result [17]. They found no deaths and only one AMI by 30 days in the 64% of patients with a negative test. The rates of positive test results are lower in chest pain centers with mandatory predischarge stress testing. Mikhail et al [18] noted that 420 patients underwent stress testing, 32 were admitted for positive results, and less than 2% received revascularization. Chan et al [11] found a revascularization rate of only 1%. Lai et al evaluated 344 patients admitted to a chest pain observation unit who did not receive stress testing before discharge and found a 60-day death or MI rate of 0.6% [12]. They concluded that low-risk patients who rule out for MI in a chest pain observation unit do not require inpatient stress testing and that the safety of discharging these patients from the chest pain observation unit is not significantly different than providing immediate stress testing [12]. This is further supported by the evidence from Mikhail et al [18] who found no diagnoses of AMI within 5 months in patients discharged form the chest pain observation unit. Chan et al [11] found no difference in 30-day outcomes regardless of whether the patients received predischarge testing, outpatient testing, or no provocative testing within the 30-day period. Although testing may help risk-stratify patients, there is no evidence that predischarge testing offers advantages over outpatient testing with regard to patient outcomes [11].

Shaver et al [19] took a different approach to studying this problem. They evaluated the impact of a negative evaluation of underlying CAD upon cardiac resource utilization over the ensuing year. They found that there was no association between a negative evaluation for underlying CAD and subsequent cardiac-related ED visits, admissions, or cardiac resource test utilization over the year after the index visit [19]. Thus, it appears that negative

Table 5 Adverse outcome by prior stress test Prior stress testing Adverse 30-day outcome, No. (%) 95% CI
Abnormal	9/89 (10.1) 3.6-16.7
None	74/1421 (5.2) 4.1-6.4
Normal	14/291 (4.8) 2.4 -7.3

tests do not provide a sufficient level of comfort to patients or providers when patients have persistent or re- current symptoms.

We extended this previous research to investigate whether knowledge of prior negative stress test impacts emergency physician clinical decision making in the ED and found that knowledge of prior negative stress test did not impact the likelihood of admission to the hospital. Admis- sion rates in patients without a prior negative test were the same as admission rates in patients that never had a test, suggesting that prior negative tests do not impact emergency physician decision making for subsequent visits.

Emergency physicians do not appear to be influenced by a prior normal stress test, but do admit patients with a prior abnormal stress test more frequently. We also found that patients with a prior normal stress test are at the same risk of adverse cardiovascular events as patients who have not previously undergone stress testing. Thus, knowledge of previously normal stress did not and should not impact clinical decision making in the ED. From a pathophysiology perspective, this makes sense. Stable angina is caused by a fixed obstruction to flow; however, unstable angina, non ST segment elevation myocardial infarction (NSTEMI) and STEMI are caused by Plaque rupture and thrombus formation in a lesion that may or may not have been significant enough to result in angina (or a positive stress test) before the acute process. Stress testing assesses whether a fixed obstruction to flow is present and cannot predict subsequent plaque rupture resulting in ischemia.

alternative methods of risk stratification may prove to be more successful in terms of both risk stratification and impact on future symptoms. deFillipi et al [20] have shown that in low-risk patients, coronary angiography is more sensitive and more cost-effective than exercise treadmill testing. Patients with negative coronary angiography had fewer repeat ED visits, fewer hospitalizations, and had higher Satisfaction rates and better understanding of their disease than patients who had negative stress tests [20]. Thus, using coronary angiography in place of stress testing low-risk patients may not only aid in patients’ understand- ing of their disease, but may also decrease ED visits and hospitalizations and may aid in physician disposition decision making. Alternatively, Computed tomography coronary angiography correlates very well with cardiac catheterization, potentially offering a noninvasive assess- ment of coronary anatomy with the same benefits on future clinical decision making and patient understanding of their disease as cardiac catheterization. Future studies should assess this possibility.

This study has several potential limitations, the first of which is selection bias. We attempted to avoid selection bias by having trained research assistants present in the ED to screen patients for study enrollment. However, these research assistants were only available 16 hours per day, so it is possible that the patients who presented during the

8 hours the research assistants were not available were

different than the patients who were enrolled. However, because the large majority of the patients (N85%) seen in the ED arrive during the time the research assistants are present, this represents only a small cohort of patients.

In addition, a patient with a prior stress test at any point in time was classified as having a prior test. Therefore, a patient who had a normal stress test 5 weeks ago was placed in the same category as a patient who had a normal stress test several years ago. It is possible that the timing of the prior normal stress test could have an impact on disposition decision making. For purposes of this analysis, all prior stress tests were considered the same because patients often do not know the type of test they received. It would require a larger number of patients with prior stress tests at different points in time to determine if timing or type of test affects physician disposition decision making.

Cardiac risk factors differed between patients who had prior abnormal, normal, or no prior stress testing. Patients with an abnormal prior stress test were significantly more likely to have hypertension, diabetes mellitus, congestive heart failure, previous MI, and to use tobacco than patients with a normal prior stress test or no prior stress test. There were significant differences between all 3 groups of patients in history of hypercholesterolemia, known CAD, and angina. Because we used revascularization as part of our secondary Composite outcome, it should be noted that this outcome might be affected by workup bias and the clinical judgment of the treating physician. It is possible, then that the difference between disposition and 30-day cardiovascu- lar outcomes may be partially explained by the differences in baseline patient characteristics as well as workup bias.

Although we were able to follow-up with 97% of our patients, we are missing data on 30-day outcome for 3% of our study population. Because this is a small fraction, however, it is not likely to greatly influence the overall results.

Lastly, our study population was derived from a single, urban, university-affiliated center, where decision making may be different than in other locations. We cannot be sure that our results generalize to other practitioners or settings.

Conclusions

In conclusion, physician knowledge of a prior stress test significantly impacts disposition decision making only in the case of an abnormal stress test. Emergency physicians do not appear to make disposition decisions based on a prior normal stress test.

References

1. McCaig LF, Burt CW. National hospital ambulatory medical care survey: 2003 emergency department summary. Advance data from vital and health statistics. Hyattsville (Md)7 National Center for Health Statistics; 2005.
2. Pope JH, Aufderheide TP, Ruthazer R, et al. missed diagnoses of Acute cardiac ischemia in the emergency department. N Engl J Med 2000;342:1163 - 70.
3. Dagnone E, Collier C, Pickett W, et al. Chest pain with nondiagnostic electrocardiogram in the emergency department: a randomized con- trolled trial of two cardiac marker regimens. CMAJ 2000;162:1561 - 6.
4. Selker HP, Beshansky JR, Griffith JL, et al. Use of the acute cardiac ischemia time-insensitive predictive instrument (ACI-TIPI) to assist with triage of patients with chest pain or other symptoms suggestive of acute cardiac ischemia. A multicenter, controlled clinical trial. Ann Intern Med 1998;129:845 - 55.
5. Limkakeng Jr A, Gibler WB, Pollack C, et al. Combination of Goldman risk and initial cardiac troponin I for emergency depart- ment chest pain patient risk stratification. Acad Emerg Med 2001; 8:696 - 702.
6. Lloyd-Jones DM, Camargo Jr CA, Lapuerta P, et al. Electrocardio- graphic and Clinical predictors of acute myocardial infarction in patients with unstable angina pectoris. Am J Cardiol 1998;81:1182 - 6.
7. Tatum JL, Jesse RL, Kontos MC, et al. Comprehensive strategy for the evaluation and triage of the chest pain patient. Ann Emerg Med 1997;29:116 - 25.
8. Reilly BM, Evans AT, Schaider JJ, et al. Impact of a clinical decision rule on hospital triage of patients with suspected acute cardiac ischemia in the emergency department. JAMA 2002;288:342 - 50.
9. Pollack Jr CV, Sites FD, Shofer FS, et al. Application of the TIMI risk score for unstable angina and non-ST elevation acute coronary syndrome to an unselected emergency department chest pain population. Acad Emerg Med 2006;13:13 - 8.
10. Antman EM, Cohen M, Bernink PJ, et al. The TIMI risk score for unstable angina/non-ST elevation MI: a method for prognostication and therapeutic decision making. JAMA 2000;284:835 - 42.
11. Chan GW, Sites FD, Shofer FS, Hollander JE. Impact of stress testing on 30-day cardiovascular outcomes for low-risk patients with

chest pain admitted to floor telemetry beds. Am J Emerg Med 2003; 21:282 - 7.

1. Lai C, Noeller TP, Schmidt K, et al. Short-term risk after initial observation for chest pain. J Emerg Med 2003;25:357 - 62.
2. Smith SW, Jackson EA, Bart BA, et al. Incidence of myocardial infarction in emergency department chest pain patients with a recent negative stress imaging test. Acad Emerg Med 2005;12:51 [abstract].
3. Hollander JE, Blomkalns AL, Brogran GX, et al. Standardized reporting guidelines for studies evaluating risk stratification of emergency department patients with potential acute coronary syn- dromes. Ann Emerg Med 2004;44:589 - 98.
4. Diercks DB, Gibler WB, Liu T, et al. Identification of patients at risk by graded exercise testing in an emergency department chest pain center. Am J Cardiol 2000;86(3):289 - 92.
5. Polanczyk CA, Johnson PA, Hartley LH, et al. Clinical correlates and prognostic significance of early negative exercise tolerance test in patients with acute chest pain seen in the hospital emergency department. Am J Cardiol 1998;81:288 - 92.
6. Amsterdam EA, Kirk JD, Diercks DB, Lewis WR, Turnipseed S. Immediate exercise testing to evaluate low risk patients presenting to the emergency department with chest pain. J Am Coll Cardiol 2002; 40:251 - 6.
7. Mikhail MG, Smith FA, Gray M, et al. Cost-effectiveness of mandatory stress testing in chest pain center patients. Ann Emerg Med 1997;29:88 - 98.
8. Shaver KJ, Marsan RJ, Sease KL, et al. Impact of a negative evaluation for underlying coronary artery disease on one-year resource utilization for patients admitted with potential acute coronary syndromes. Acad Emerg Med 2004;11:1272 - 7.
9. deFilippi CR, Rosanio S, Tocchi M, et al. Randomization of a strategy of predischarge coronary angiography versus exercise testing in low- risk patients in a chest pain unit: in-hospital and long-term outcomes. J Am Coll Cardiol 2001;37:2042 - 9.