Evidence Report/Technology Assessment

Number 31

Prepared for:
Agency for Healthcare Research and Quality

U.S. Department of Health and Human Services
2101 East Jefferson Street
Rockville, MD 20852
http://www.ahrq.gov

Contract No. 290-97-0013

Prepared by:
UCSF-Stanford Evidence-based Practice Center
Paul A. Heidenreich, MD, MS
Principal Investigator

Alan Go, MD, MPH
Kathryn A. Melsop, MS
Thomas Alloggiamento, MD
Kathryn M. McDonald, MM
Vivian Hagan
Trevor Hastie, PhD
Mark A. Hlatky, MD
Investigators

AHRQ Publication No. 01-E001

December 2000

On December 6, 1999, under Public Law 106-129, the Agency for Health Care Policy and Research (AHCPR) was reauthorized and renamed the Agency for Healthcare Research and Quality (AHRQ). The law authorizes AHRQ to continue its research on the cost, quality, and outcomes of health care and expands its role to improve patient safety and address medical errors.

This report may be used, in whole or in part, as the basis for development of clinical practice guidelines and other quality enhancement tools, or a basis for reimbursement and coverage policies. AHRQ or U.S. Department of Health and Human Services endorsement of such derivative products may not be stated or implied.

ISBN 1-58763-012-5

ISSN 1530-4396

The Agency for Healthcare Research and Quality (AHRQ), formerly the Agency for Health Care Policy and Research, through its Evidence-based Practice Centers (EPCs), sponsors the development of evidence reports and technology assessments to assist public and private-sector organizations in their efforts to improve the quality of health care in the United States. The reports and assessments provide organizations with comprehensive, science-based information on common, costly medical conditions and new health care technologies. The EPCs systematically review the relevant scientific literature on topics assigned to them by AHRQ and conduct additional analyses when appropriate prior to developing their reports and assessments.

To bring the broadest range of experts into the development of evidence reports and health technology assessments, AHRQ encourages the EPCs to form partnerships and enter into collaborations with other medical and research organizations. The EPCs work with these partner organizations to ensure that the evidence reports and technology assessments they produce will become building blocks for health care quality improvement projects throughout the Nation. The reports undergo peer review prior to their release.

AHRQ expects that the EPC evidence reports and technology assessments will inform individual health plans, providers, and purchasers as well as the health care system as a whole by providing important information to help improve health care quality.

We welcome written comments on this evidence report. They may be sent to: Director, Center for Practice and Technology Assessment, Agency for Healthcare Research and Quality, 6010 Executive Blvd., Suite 300, Rockville, MD 20852.

John M. Eisenberg, M.D.	Douglas B. Kamerow, M.D.
Director Agency for Healthcare Research and Quality	Director, Center for Practice and Technology Assessment Agency for Healthcare Research and Quality

The authors of this report are responsible for its content. Statements in the report should not be construed as endorsement by the Agency for Healthcare Research and Quality or the U.S. Department of Health and Human Services of a particular drug, test, treatment, or other clinical service.

Objective. Unstable angina comprises a broad spectrum of ischemic heart disease and is associated with varying levels of risk for unfavorable outcomes including myocardial infarction and death. Despite development of various diagnostic approaches, the evaluation of patients with chest pain suggestive of unstable angina or myocardial infarction remains a common, costly problem, with approximately 5 million people undergoing evaluation in emergency departments annually at an estimated cost of over $6 billion. The American College of Cardiology and the American Heart Association established a committee to develop guidelines for the diagnosis and treatment of unstable angina. Under a contract with the Agency for Health Care Policy and Research to assist the committee to evaluate the current ability to predict risk for patients with unstable angina, we performed three systematic reviews. The first review concerned the value of the electrocardiogram, physical examination, and clinical history in predicting outcome for patients with unstable angina. The second review examined the ability of troponin to predict outcome in patients with proven or suspected unstable angina. The third review examined the efficacy of chest pain units and emergency department protocols in patients who have suspected unstable angina or myocardial infarction.

Search Strategy. We identified published studies (English language) through 1998 by searching the MEDLINE and EMBASE databases and by manually reviewing the bibliographies of identified articles.

Selection Criteria. For the review of clinical and electrocardiographic predictors of outcome, we restricted our review to only those studies that performed a multivariate analysis of the clinical and/or electrocardiographic predictors of adverse clinical events in patients with either chest pain suggestive of ischemia or diagnosed unstable angina in the emergency department or hospital. For the review of troponin efficacy, we included reports of patient cohorts with unstable angina or suspected unstable angina that noted subsequent myocardial infarction, death, or revascularization. For the review of chest pain units and emergency department protocols, we included trials that were randomized. We also included controlled clinical trials of chest pain protocols used in the emergency department.

Data Collection and Analysis. For the review of troponin studies, we pooled the data using odds ratios and relative risks for outcomes of death, subsequent myocardial infarction, and revascularization. Two independent reviewers abstracted each study.

Main Results. Characteristics of patients with suspected unstable angina that were associated with worse outcomes included advanced age, male sex, prior myocardial infarction, and diabetes. In addition, congestive heart failure, hypertension, and smoking may also be important prognostic factors from the clinical history, but specific descriptors of the chest pain did not provide prognostic information. The strongest electrocardiographic predictor of adverse outcomes was ST-segment depression greater than 0.1 millivolt, whereas a completely normal electrocardiogram was a strong predictor of reduced risk. A positive troponin finding increased the risk of subsequent death 5.3-fold at 4 weeks (95 percent confidence interval: 3.6-7.9). A positive troponin finding also increased the risk of subsequent death or myocardial infarction 12.3-fold at 4 weeks (95 percent confidence interval: 6.4-23.8) in patients with diagnosed unstable angina. The absolute increase in mortality was 3.9 percent (95 percent confidence interval: 3.0-4.4) for patients with a positive troponin level. The predictive value of troponin T and troponin I was not significantly different. Data are insufficient at present to determine if rapid bedside troponin tests and laboratory-based measurements provide similar or different prognostic information. Although randomized trials of chest pain units are few, they consistently have shown decreased hospital costs compared with usual care. All studies included in this report apply to adult men and women.

Conclusions. Several patient characteristics and electrocardiographic findings portend a worse prognosis in patients with suspected or diagnosed unstable angina including older age, male sex, past myocardial infarction, diabetes mellitus, and ST depression greater than 0.1 millivolt. Measurement of troponin T or troponin I provides additional independent prognostic information. Additional randomized trials of chest pain units are needed to determine more fully their health and economic benefits.

This document is in the public domain and may be used and reprinted without permission except those copyrighted materials noted for which further reproduction is prohibited without the specific permission of copyright holders.

Suggested Citation:
Heidenreich PA, Go A, Melsop KA, et al. Prediction of risk for patients with unstable angina. Evidence Report/Technology Assessment No. 31 (prepared by the UCSF-Stanford Evidence-based Practice Center under Contract No. 290-97-0013). AHRQ Publication No. 01-E001. Rockville, MD: Agency for Healthcare Research and Quality. December 2000.

Coronary heart disease is the leading cause of death for both men and women in the United States. One of the most characteristic and troubling features of coronary disease is the sudden and unexpected onset of symptoms in clinically stable patients and sometimes in even previously healthy individuals.

The development of symptoms is associated with an increased risk of sudden death, acute myocardial infarction, and other life-threatening complications. The development of symptoms suggestive of coronary disease, therefore, mandates prompt and accurate diagnosis and treatment.

The cardinal symptom of coronary artery disease (CAD) is angina, which classically presents as a squeezing or strangulating deep chest discomfort that may radiate to the arm or jaw. Angina that is brought on by exercise stress and is relieved promptly after cessation of exertion is termed "typical angina." Stable angina is a pattern of symptoms that has been unchanged for 6 or more weeks. Unstable angina is a pattern of symptoms that is new in onset, changing in severity or frequency, occurring at rest, or lasting longer than 20 minutes.

The evaluation of suspected coronary disease is complicated by the fact that chest discomfort has many causes, and bona fide coronary disease may present in an atypical fashion. Thus, a population of patients with symptoms suggestive of coronary disease includes some patients with acute, life-threatening medical problems, some patients with other medical problems mimicking CAD, and even some "worried well" in need only of reassurance. The evaluation and treatment of this highly heterogeneous population is the difficult task for clinicians in emergency departments (EDs) and in office practice. The key goal of these clinicians must be to identify the patient's short-term risk. The high-risk patient may develop life-threatening complications and require hospitalization and immediate therapy. The low-risk patient may need further evaluation, but in a less urgent and less costly setting. Because identification of patient risk is central to all further patient management in unstable angina, this evidence report focuses on clinical and laboratory markers of patient risk, such as results of diagnostic tests (troponin values, stress testing, echocardiography, and nuclear scintigraphy). Because chest pain units attempt to "risk stratify" (group patients according to their degree of risk) based on readily available data, an assessment of the efficacy of the chest pain units is significant to this report. Our in-depth review focused on information that would be readily available to all providers caring for patients with suspected unstable angina. Information in this evidence report applies to adult men and women.

We conducted a systematic literature search of MEDLINE for relevant articles published between 1966 through 1998, and we manually searched references of retrieved articles to identify additional applicable published studies. Our search criteria included MEDLINE exact subject and keyword searches for:

• Chest pain, angina pectoris, unstable angina, variant angina, vasospastic angina, or acute coronary syndrome combined with the terms risk, stratification, prognosis, outcome, and multivariate analysis.

We restricted our review only to those studies that performed a multivariate analysis of the clinical and/or electrocardiographic predictors of adverse clinical events in patients with either chest pain suspected to be ischemia or diagnosed unstable angina in the ED or hospital. We sought to determine those variables that provided independent risk prediction. Therefore, studies were excluded if they performed a multivariate analysis but did not provide the quantitative results with measures of significance (e.g., p value, standard error) in the form of regression estimates, relative risks, odds ratios, or rate ratios. A p value of <0.05 was considered statistically significant. We also excluded non-English language studies.

Predictor variables of interest included:

• Demographic characteristics (age, sex, race/ethnicity).
• Medical history (prior MI, unstable or stable angina, revascularization, congestive heart failure, cerebrovascular disease, hypertension, diabetes, and smoking history).
• Symptom characteristics (frequency, duration, and pattern of chest pain).
• Initial physical exam findings (blood pressure, heart rate, and pulmonary rales or Killip class I-IV for congestive heart failure).
• Initial electrocardiographic features (ST-segment depression, transient ST elevation, isolated T-wave inversions, other findings, or a normal electrocardiogram).

We included studies that measured at least one of the following outcomes:

• Cardiac death.
• Myocardial infarction.
• Other major cardiac complication.

Candidate titles and abstracts were reviewed, and appropriate studies were selected for data extraction by an internist and a cardiologist with training in health services research.

Studies were stratified by type of patients evaluated (chest pain or diagnosed unstable angina), and multivariate results were grouped into:

• Categories of demographic characteristics.
• Medical history features.
• Symptom characteristics.
• Initial physical findings.
• Electrocardiographic features.

We searched MEDLINE (1966-98) and EMBASE (1974-98) and reviewed cited references of retrieved articles to identify relevant published studies. Our search criteria were: (1) the text word troponin, (2) the text words angina or unstable or myocardial infarction or ischemia, and (3) language English, excluding (4) the MeSH heading animal. Using this criteria set, we searched MEDLINE and then EMBASE. Finally, we reviewed the bibliographies of identified trials to locate other relevant studies.

We restricted our review to studies that evaluated patient cohorts with suspected ischemia. We excluded studies that only enrolled patients with myocardial infarction. We also excluded case-control studies and studies that did not report the outcomes of MI or death.

Study selection was performed initially by title review. Candidate abstracts were then reviewed and selected for data abstraction. Two independent reviewers abstracted data from each article on standardized electronic data forms. A third reviewer compared their results and settled any differences. In general, at least one reviewer of the pair had clinical cardiology expertise.

We used standard methods of meta-analysis to combine outcome data across trials and the Peto (fixed-effects) and DerSimonian-Laird (random-effects) methods to estimate summary odds ratios. We examined differences between study subgroups using analysis of variance. The previously specified subgroup comparisons of interest were studies for all patients with suspected ischemia versus patients in whom myocardial infarction had already been excluded. The former patients would be recruited from emergency departments, while the latter patients would all be hospitalized. Reported p values are two tailed with statistical significance at p<0.05.

We conducted a systematic literature search of MEDLINE for articles published between 1966 and 1998.

Abstracts and titles from both searches were reviewed for appropriate studies. Randomized trials or controlled clinical trials were selected. Studies that assessed chest pain units, accelerated or rapid diagnostic protocols, or emergency department triage protocols were reviewed. Noncontrolled studies were selected if they reported outcomes for at least 1,000 patients with suspected acute coronary syndromes. We recorded the outcomes of hospital admission rate, cost of care, myocardial infarction, and death when available. Other outcomes reported by the trial were recorded if comparisons were made between control and intervention groups.

• Demographic characteristics associated with worse outcomes included increasing age and male sex.
• Prior medical conditions that consistently predicted poor outcomes included previous myocardial infarction and diabetes. In addition, prior congestive heart failure, hypertension, and smoking also may be important prognostic factors.
• Congestive heart failure on presentation increased the risk of cardiac events.
• Among patients with suspected or confirmed unstable angina, specific characteristics of the chest pain did not uniformly add any independent information useful in predicting adverse outcomes among the studies reviewed.
• ST depression >0.1 mV was the strongest electrocardiographic predictor of adverse outcomes, while a completely normal electrocardiogram was a strong predictor of reduced risk.

• A positive troponin value increased the risk of death 5.3-fold over the 4 weeks subsequent to testing (95 percent confidence interval [CI]: 3.6-7.9).
• The absolute increase in mortality was 3.9 percent (95 percent CI: 3.0-4.4) for patients with a positive troponin level.
• A positive troponin value increased the risk of subsequent death or MI 12.3-fold at 4 weeks (95 percent CI: 6.4-23.8).
• The absolute increase in the rate of death or subsequent MI was 14 percent (95 percent CI: 10-18) for patients with a positive troponin level.
• The increased absolute risk associated with a positive troponin value was proportional to the overall risk of death.
• The increased risk of death associated with a positive troponin value decreased over time.
• The increased risk of death at 4 weeks was similar for elevated troponin T and troponin I values.
• An increased troponin value was associated with a similar risk of death for both patients with unstable angina (MI excluded) and patients with suspected acute ischemia (chest pain with or without MI).

• The few randomized trials of chest pain units have consistently shown decreased hospital days and hospital costs for patients as compared with usual emergency care for the duration of the initial encounter.
• There is no evidence of increased harm from the more selective admissions that result from evaluation with chest pain units or protocols, but statistical power is limited in studies to date.
• Patient satisfaction may be improved by chest pain units.

More studies are needed that combine data from history, physical examination, and biochemical markers to determine the independent prognostic ability of each variable. Our qualitative review of clinical and electrocardiographic predictors of prognosis suggests that future studies should include the following variables in a multivariate prediction model:

• Age.
• ST depression.
• ST elevation.
• Troponin levels.

More randomized trials of chest pain units or chest pain protocols are needed to more fully determine their health and economic benefits.