|
|||||||
|
|
|
|||||
|
|
|||||||
|
|||||||
|
|
|
|||||
|
|
|||||||
Author: Debra R. Judelson, MD, FACC, FACP, Cardiovascular Medical Group of Southern California, Women's Cardiovascular Institute of Southern California
Abstract: Gender discrepancies have developed in the evaluation of coronary heart disease (CHD), arising from such early myths as "CHD is a man's disease." The challenge is to make sure that the noninvasive testing for CHD in women is sensitive and specific enough to lead to the correct treatment. Coronary angiography, the gold standard for CHD diagnosis, must be interpreted along with functional information. The standard noninvasive test--stress electrocardiograph (ECG)--is associated with up to 40% false-positive S-T segment depressions in women, versus fewer than 10% in men. The predictive value of exercise stress testing in women is particularly poor. In one study, stress ECG had a specificity of 61%, a sensitivity of 68%, a positive predictive value of 0.61, and a negative predictive value of 0.68. Stress echocardiography can have high sensitivity (86%) and specificity (86%), but often examiners stop the test before detecting less severe areas of damage. Also, acquiring adequate images is difficult in women with breast implants or large breasts. Nuclear perfusion imaging with thallium-201 has shown a sensitivity of 84% to 90% and a specificity of 75% to 87% in women, but the diagnostic accuracy can be reduced in patients who are obese or have large breasts. A higher-energy radiotracer, technetium-99m (Tc-99m) sestamibi, has been introduced. In one study, the sensitivity of the 2 agents was similar (85% to 90%), while the specificity of Tc-99m was higher (84% to 94%) than that of thallium-201 (71%). [Medscape Women's Health 2(2), 1997. © 1997 Medscape, Inc.]
Introduction
Coronary heart disease (CHD) in women has received much attention in the media and in medical literature because of its recent recognition as the primary cause of death in American women. Articles have addressed perceived and real gender bias in the way CHD has been studied and treated in women. Much of the focus has been on how gender differences lead to discrepancies in how the medical community evaluates risk factors, as well as how symptoms are interpreted and managed. Women with CHD often experience delays in treatment and referral for cardiac catheterization and must overcome obstacles to revascularization and cardiac rehabilitation. As these issues are in the process of being resolved, the problem of evaluation and diagnostic testing for CHD in women remains an area that must be specifically addressed, because improved understanding will have an immediate impact on women's health care. In this article, we concentrate on issues in evaluation and diagnostic testing for CHD in women in the outpatient or non-emergent setting.
Gender Discrepancies in CHD Evaluation
Gender discrepancies exist in the evaluation of CHD evaluation of women for several reasons. The most prominent is the myth perpetuated by the Framingham Study and other studies that CHD is primarily a man's disease. In the Framingham Study, chest pain was interpreted as being synonymous with CHD at a time when angiographic confirmation was not available. At 6 years' follow-up, none of the women initially presenting with chest pain sustained a myocardial infarction (MI). At 12 years' follow-up, 70% of the men who initially had chest pain had died of cardiac disease, while only 31% of the women who initially had chest pain died of cardiac disease. These data led to decades of physicians believing that chest pain was generally benign in women.[1]
The Coronary Artery Surgery Study (CASS) Registry[2] further demonstrated how poorly chest pain in women correlated to CHD. In patients referred for cardiac catheterization to evaluate chest pain, a gender comparison of the incidence of significant coronary artery disease was performed. In that study, 50% of women with chest pain showed little or no coronary artery disease on angiogram, while only 17% of men tested had little or no coronary artery disease. This gender difference was seen regardless of which coronary risk factors were present. This study forced the realization that while chest pain is a major indicator of CHD in men, a history of chest pain alone is a poor predictor of CHD in women.
Yet CHD is common in women, reportedly occurring in 1 in 9 women over the age of 45 years and 1 in 3 women over the age of 65 years, according to data from the American Heart Association. Of the 250,000 women who die annually from CHD, 60% are over the age of 65 years. The late age of CHD presentation in most women relative to men, as well as its less typical presentation, may contribute to the underappreciation of women's heart disease.
The misperception held by the public and by the medical profession that CHD is not a woman's problem continues to this day, despite increasing evidence to the contrary. A 1995 Gallup Survey indicated that 80% of women aged 45 to 75 years and 32% of primary care physicians did not know that heart disease was the number one cause of death in American women. This lack of knowledge about cardiac risk leads to the double problem of women not seeking medical evaluation for potential cardiac symptoms and physicians not pursuing appropriate diagnostic testing.
The male model of CHD that is taught to physicians also leads to the underestimation of cardiac risk in women. Further, many physicians fail to recognize diabetes and smoking as more significant risk factors for CHD in women than in men. Lipid risk evaluation and treatment are traditionally based on LDL abnormalities, which are of primary importance for men, whereas HDL and triglyceride levels, which are more important indicators in women, are barely addressed. The dramatic impact of menopause, which causes a 20-fold increase in the incidence of heart disease, was not appreciated by the majority of primary care physicians polled in the Gallup Survey.
The medical profession's view of cardiac ischemia is focused on the younger, active male and his type of ischemic symptoms. The woman with cardiac disease is likely to be 10 years older than the typical man, an age when other complications such as diabetes and hypertension tend to emerge or escalate; hypertension can raise the risk of MI, while older age and diabetes can blunt the woman's ability to feel the classic sign of ischemia--pain. In addition, the "typical" middle-aged woman with cardiac disease is less likely to be physically active. Although the complaint of substernal chest pain or pressure coming on with physical activity and diminishing with rest is the same for men and women with CHD, a woman with CHD is also likely to experience ischemic symptoms of dyspnea, gastrointestinal symptoms, or weakness during emotional stress or at rest. This "atypical" presentation often causes physicians to rule out ischemia and, therefore, is often not followed up with appropriate diagnostic testing, regardless of a woman's true risk.
Gender discrepancies in evaluation are compounded by a reliance on diagnostic testing procedures that have not been validated in women. When diagnostic testing has unacceptable levels of false-positive and false-negative test results, there is a legitimate lack of confidence in the results, which prevents the timely referral of women with abnormal findings for invasive testing and revascularization, leading to excessive morbidity and mortality from CHD.
Background to Noninvasive Diagnostic Testing for CHD
In order to understand issues involved in gender-sensitive testing, physicians must understand the nature of noninvasive testing for CHD and be clear on the goal: diagnosis of CHD by noninvasive testing when the patient's health is stable permits symptoms to be treated by lifestyle changes and medication and provides a window of time for consideration of referral for invasive testing and revascularization, if appropriate. CHD is a progressive atherosclerotic process of gradual narrowing of the coronary vessels, leading to ischemia when the myocardium served by the vessel requires more blood flow than it receives.
Acute MI, on the other hand, usually occurs when there is a sudden rupture of an atherosclerotic plaque that causes obstruction of the coronary vessel, creating an unstable situation requiring immediate inpatient intervention. The blood vessels with severe stenoses are not necessarily the arteries that cause acute MIs.[3,4] Since physicians do not have the technology to easily identify which plaques will rupture, our noninvasive diagnostic testing has been concentrated on identifying coronary-vessel narrowing that causes angina and making treatment decisions based on those results.
Our interest in diagnosing narrowing of coronary blood vessels is focused on identifying the fixed, flow-limiting disease that is causing our patients' symptoms. Many cardiologists believe that proceeding to cardiac catheterization any time there is concern is the appropriate route to take. Even if the costs and risks to the patient are not considered, angiography provides only part of the information necessary to make a decision for the patient. Angiography tells us about the stenoses present, but their significance is not particularly apparent unless more than 90% of all lesions are visualized. For example, the presence of multiple lesions less than 50% occluded, while not independently viewed as flow-limiting, may significantly impede myocardial perfusion. An artery with 70% stenoses, usually considered as limiting blood flow, may be under- or overestimated, especially in patients with diffuse disease or small coronary arteries. Collateral vessels serving as a patient's own bypass may be present, but their full impact may not be appreciated. In patients with prior MI, coronary lesions may serve myocardium that is either viable or nonviable. Therefore, identification of narrowing is not sufficient to direct the thoughtful physician.
While still viewed as definitive for diagnosis of coronary artery disease, coronary angiography findings must be interpreted along with functional information to determine the significance of coronary lesions in the stable patient, so that revascularization decisions can be effectively made. In the evaluation of a patient with possible angina, the determination of functional information is helpful as a first step. If reliable studies are negative--indicating the absence of significant coronary artery narrowing--angiography can be deferred and the patient reassured. If only minimal amounts of myocardium are identified as being at risk in reliable studies, medical therapy may be appropriate without further delineation of anatomy. In these patients, the annual mortality from CHD is less than 1%, too low to warrant coronary revascularization, which may be associated with greater risk and is unlikely to increase survival any more than would medical therapy alone.[5] However, if noninvasive studies show enough myocardium at risk to benefit from revascularization, or if symptoms cannot be controlled, invasive testing is indicated. Our challenge is to make sure that the noninvasive testing for CHD in women is sensitive and specific enough to lead us to the correct treatment route.
Standard noninvasive testing uses physical exercise or stress to identify fixed narrowing of coronary arteries. The discrepancy between myocardial supply and demand in patients with coronary narrowing becomes apparent if the stress level is high enough to make the heart work harder. Maximal exercise on a bicycle or treadmill causes an increase in blood pressure and cardiac work along with a 2-fold increase in normal coronary blood flow, which is not seen in flow-restricted vessels. Noninvasive tests identify either the variation in blood flow or the resultant ischemia.[6]
In patients who are unable to exercise, pharmacologic means are used to mimic the reaction of the body to stress. Such drugs as dipyridamole or adenosine may be used to achieve a hyperemic effect--that is, an increase in coronary blood flow without affecting cardiac muscle contraction. Dipyridamole causes a mild 5- to 10-beat increase in heart rate and a slight decrease in arterial blood pressure. There is a 4- to 6-fold increase in normal coronary blood flow; the vascular bed distal to significant (>50%) coronary lesions is already maximally dilated and does not dilate further with dipyridamole. Ischemia does not need to be induced to demonstrate this blood flow discrepancy in regional perfusion. Adenosine has a similar action, though it has a more rapid onset and wears off faster. It is associated with a higher rate of heart block, though this is transient. Both drugs can cause brief neurologic abnormalities or shortness of breath, which is reversible with theophylline. The antagonistic effect of theophylline, however, means that neither dipyridamole nor adenosine can be used in a pharmacologic stress test of patients who are on theophylline.
Pharmacologic means can also be used to raise cardiac work with only a slight increase in blood flow. The prototype drug used is dobutamine. Low-dose dobutamine increases the rate and force of contractility with mild peripheral vasodilatation. High-dose dobutamine causes a significant increase in heart rate, contractility, and blood pressure; any resulting ischemia can be identified with noninvasive testing. Major side effects of dobutamine include chest pain and arrhythmias; however, the medication is usually well tolerated.
To reach diagnostic levels of stress in noninvasive evaluations of coronary function, pharmacologic stress tests are required in 55% to 60% of women, but only 40% of men; in contrast, exercise stress tests are used in only 40% to 45% of women and in 60% of men. These testing trends address the lower exercise capability of the typically older, less active female cardiac patient. Without the use of pharmacologic stress in patients unable to reach maximal exercise levels, noninvasive testing has unacceptable rates of false-negative results and prevents timely referral of patients for evaluation by invasive means.
Specific Noninvasive Diagnostic Tests
Electrocardiograph (ECG) stress test. The standard noninvasive test is the exercise, or stress, electrocardiograph (ECG) test. This test is widely available and relatively inexpensive; however, the sensitivity and specificity of ECG monitoring are relatively low. We usually identify S-T segment depressions on continuous ECG monitoring as indicative of coronary ischemia, but the diagnostic value of S-T segment depression has not been validated in women. Multiple studies[7-11] have demonstrated the diagnostic reliability of exercise stress testing to be as low as 52%, with up to 40% false-positive S-T segment depressions in women, versus fewer than 10% in men. The predictive value of exercise stress testing in women is particularly poor. In a recent study of women referred for outpatient testing and catheterization for evaluation of chest pain,[12] the stress ECG had a specificity of 61%, a sensitivity of 68%, a positive predictive value of 0.61, and a negative predictive value of 0.68 (see Table I). Results were false-positive or false-negative in 36% of the women. However, the predictive value of a negative test was higher in younger women and in those with few cardiac risk factors than in older women or those with multiple cardiac risk factors.
The recurrent finding of an abnormal exercise ECG in women without corresponding coronary artery disease on angiography has led to a lack of confidence in the diagnostic and prognostic value of the exercise ECG test. Physicians accustomed to being misled by abnormalities will often fail to refer patients for further testing, regardless of the risks identified, thus adding to the diagnostic dilemma for women. The stress ECG test is best reserved for use in younger women at low risk for CHD in whom the negative predictive value is greater. In these younger patients, the demonstrated ability to exercise safely is of greater importance than the test's reliability in identifying coronary artery disease. Yet even in this group, if abnormalities are discovered, referral for more sensitive testing is still needed.
Stress echocardiography. Other noninvasive tests have been developed in response to the limitations of stress ECG. For example, stress or exercise echocardiography is a popular noninvasive diagnostic test for several reasons: It is convenient and often available in the local cardiologist's office, avoids ionizing radiation, and can identify cardiac structure and left ventricular function. To be diagnostic of coronary narrowing, exercise echo requires the development of ischemia with resultant regional wall motion abnormalities. Since more severely narrowed areas show wall motion abnormalities first, care must be taken to continue exercise to a maximal level if the full degree of myocardium at risk is to be identified. This is often a problem, as many examiners stop the exercise testing once wall motion abnormalities are seen. While this test can effectively confirm the presence of coronary disease, it is not as helpful in aiding the referral to catheterization and revascularization unless maximal testing is done.
When performed properly in female populations at high risk, stress echo has high sensitivity (86%) and specificity (86%) for CHD.[13] In more representative populations including both males and females, stress echo has demonstrated an average sensitivity of 80% (range, 70% to 93%) and an average-to-low specificity of 37% to 96% confirmed by angiography.[14-17] In patients who are unable to exercise, pharmacologically induced stress is extremely useful with echocardiography. Dobutamine has been found to have higher sensitivity and specificity than dipyridamole and adenosine at regular doses in stress echo.[18] Recent studies with high-dose adenosine have shown higher sensitivity (92%) and specificity (88%) than the normal doses, as well as improvement in single- and multiple-vessel disease identification; however, these results were not differentiated based on gender.[19]
Calculation of ejection fraction at rest and a demonstrated increase at peak exercise can support the diagnosis of a normal stress echo study. Unfortunately, a true difference between women and men has been identified that makes this portion of the test unreliable for women. A normal response to exercise in men is to increase their ejection fraction by 5% or more. In women, smaller resting ventricular volumes and an increase in left ventricular end-diastolic volume cause an increase in exercise stroke volume without an increase in ejection fraction. Therefore, the corroborating evidence often required for a "normal" study is lacking in women.[20]
Another disadvantage associated with stress echo is the difficulty of acquiring adequate images in 10% to 15% of all patients because of such such factors as obesity, pulmonary disease, breast implants, and large breasts. This test is very operator dependent, its interpretation is subjective and nonstandardized, and considerable expertise on the part of the interpreting cardiologist is required. If upright treadmill exercise testing is performed, peak exercise images must be obtained after exercise, reducing the sensitivity of the test because ischemic abnormalities may resolve quickly. This problem can be diminished by using supine exercise with continuous echocardiographic imaging or pharmacologically induced stress.[21]
Because of the universal availability of echocardiography in cardiologists' offices, stress echocardiography has become widely available without being subjected to the same strict standards that are required in major medical and research centers. The results of this excellent test must be correlated with angiographic findings to make sure that its reliability in local offices is as good as that seen in tertiary care centers where sensitivity and specificity determinations have been made.
Nuclear perfusion imaging. Exercise myocardial perfusion imaging has been widely available for noninvasive diagnosis of coronary artery disease. Unlike stress echocardiography, nuclear imaging is usually confined to medical centers with specially trained nuclear radiologists and cardiologists, using computerized acquisition systems that reduce operator variability. The traditional nuclear imaging agent is thallium-201. Its uptake by the myocardium is proportional to regional myocardial blood flow and viable myocardial cells. In addition, ischemic myocardial cells will not take up thallium-201, creating an image defect. Thallium-201 is injected at peak exercise, and stress images are rapidly obtained. Redistribution images are obtained at least 4 hours later. The difference between the stress and redistribution images demonstrates areas of reversible ischemia indicating coronary artery disease. Defects can be obtained with regional blood flow discrepancies even without the development of ischemia. Nuclear perfusion imaging with thallium-201 is comparable to stress echocardiography in terms of sensitivity and specificity, with results from one study showing a sensitivity of 84% to 90% and specificity of 75% to 87% in women.[22]
The diagnostic accuracy of thallium-201 perfusion imaging is reduced in obese patients because of scatter and attenuation that obscures the inferior wall of the heart and in patients with large breasts because artifact is created that obscures the anterior wall of the heart. Several models have been developed to reduce the effect of the breast artifact; however, diagnostically inadequate results occur in 11% of women and 2% of men.[23] These problem areas lead to overdiagnosis of CHD by misinterpreting artifact as blood flow discrepancies, or underdiagnosis of CHD by masking true defects. If peak imaging is not obtained promptly at the cessation of exercise, false-negative results will be obtained because of the rapid redistribution of thallium-201. This is a significant problem in medical centers where the exercise study is not performed in the nuclear medicine department, where images are acquired.
A higher-energy radiotracer, technetium-99m (Tc-99m) sestamibi (Figures. 1 and 2), is also used to obtain nuclear perfusion imaging studies in a similar fashion as thallium-201. Regional blood flow disparities are demonstrated with the same gamma nuclear counters as thallium-201 uses. Tc-99m sestamibi differs from thallium-201 in that it has less scatter and attenuation, thus reducing artifacts from obesity and breast tissue. Tc-99m sestamibi takes a snapshot of the myocardium perfusion at peak exercise, freezing the images of perfusion abnormalities. This is advantageous because immediate imaging is not necessary after exercise, allowing for greater flexibility in scheduling and re-imaging, especially to recheck for possible artifacts. A disadvantage of this feature is that Tc-99m requires a second injection to obtain the at-rest images. Tc-99m sestamibi can also be used to evaluate wall motion and ejection fraction using the same injection as the myocardial perfusion stress test. Including this technique, called ECG-gated single-photon emission computed tomographic (SPECT) imaging, improves overall test reliability even further.
The use of Tc-99m sestamibi to diagnose CHD in women is associated with comparable or higher sensitivity and specificity compared with thallium-201. In a study of 115 women, the sensitivity of both agents was similar (84% to 90%), while the specificity of Tc-99m sestamibi was higher (84% without ECG-gated SPECT, 94% with ECG-gated SPECT) than that of thallium-201 (71%).[24] Tc-99m sestamibi nuclear testing provides excellent prognostic information, especially in women, and can be combined with thallium-201 scanning at rest (dual-isotope imaging). Risk stratification in recent studies showed higher event rates in women with abnormal scans (11.5%) than in men with abnormal scans (5.8%).[25]
Clinical Decision-Making in Noninvasive Diagnostic Testing
Noninvasive testing for CHD in women is used to evaluate chest pain syndromes as well as to evaluate patients at risk for CHD, regardless of symptoms. Because of the prevalence of CHD in women and the problems of under-recognition and failure to refer women for invasive testing in a timely manner, special attention should be given to ensure referral and evaluation of appropriate patients for revascularization. A greater increase in mortality and morbidity in women following MI and bypass surgery than in men,[26] coupled with cost constraints, forces us to make more thoughtful decisions about testing. Our diagnostic tests must address the appropriate questions: Does this patient have coronary artery disease and is she likely to benefit from invasive evaluation and intervention?
Physicians responsible for evaluating women at risk for CHD must make several decisions based on clinical factors. An atypical quality or description of a patient's symptoms should not deter the physician from giving an appropriate evaluation of a female patient, especially if she has multiple risk factors. The selection of a noninvasive diagnostic test includes consideration of patient risk factors, the patient's ability to exercise, the quality of particular techniques in your community, cost, and physician/patient convenience.
Low-risk patients with normal resting ECGs, especially young and physically active women, may be safely evaluated with a maximal stress ECG test. A negative maximal test in this situation has a good predictive value. This recommendation does not hold for high-risk women in whom the predictive value of a negative test is unacceptably low. These patients should be evaluated with an imaging technique to provide reasonable sensitivity and specificity upon which treatment decisions and referral to coronary angiography can be based.
Failure to obtain a maximal stress test is the most common cause of a false-negative result that misses the diagnosis of CHD. In patients unable to exercise maximally, pharmacologic stress agents should be used to provide a diagnostic test result. The choice of pharmacologic agent depends on the imaging technique used and the preferences of the testing facility and physicians. Recent studies demonstrate similar results with the available agents in skilled situations.
The choice of imaging technique for patients at higher risk should be made carefully, with full knowledge of the level of expertise in a particular testing center. Published articles have reported on the sensitivity and specificity of imaging techniques documented at research centers, not those in local community practices. Therefore, common sense dictates that physicians be aware of how the results of their communities' noninvasive diagnostic testing compare with those of coronary angiography. In order to determine which tests are preferable, physicians must have expertise in interpreting these results.
The ease and wide availability of stress echocardiography, as well as its lack of ionizing radiation exposure, make this a desirable imaging test for women. However, the lack of standardization and subjective interpretation in stress echocardiography mean that a lower sensitivity and specificity may be found in local community practices than in major research centers. If a community has excellent correlation of stress echocardiography with angiographic results, then stress echocardiography can be used reliably. If the correlation is less precise, the more rigorous nuclear perfusion imaging will provide patients and their physicians with greater reliability of testing and better patient outcome.
Nuclear perfusion imaging requires the more costly equipment usually found at major hospital and research centers, but it provides greater standardization and more rigorous interpretation by specialized nuclear radiologists and cardiologists. In centers with SPECT capability, the test results in practice are similar to those obtained in research studies.
The cost of each test includes physician charges as well as facility and material costs. Tremendous variability exists in pricing, and physicians should be familiar with the costs under various contractual plans for each test in their community. The need to do multiple follow-up tests if nondiagnostic results are obtained may unnecessarily increase the cost of testing, so that following the inexpensive stress ECG with a more costly test becomes more expensive than doing the correct test initially. These calculations do not even take into account the hidden cost of a missed diagnosis of heart disease or the personal and emotional cost to a patient of a mistaken diagnosis of heart disease.
Recent studies show that risk stratification with nuclear perfusion imaging provides a unique opportunity to optimize testing in high-risk women. Dual-isotope myocardial perfusion imaging provides superior stratification for evaluating women at high risk for CHD and may be the most cost-effective way to evaluate this group of patients, rapidly and accurately steering the patients at greatest risk for future cardiac complications to prompt invasive evaluation.
Conclusion
The functional information provided by appropriate noninvasive cardiac diagnostic testing is useful in determining which patients can be safely reassured, which can be safely treated medically, and which should be directed to invasive evaluation and intervention. This is particularly relevant for women, in whom the presentation of cardiac symptoms and risks for CHD may be underappreciated. Noninvasive diagnostic testing for CHD in women must take into account the sensitivity and specificity of the chosen method in women, as well as the actual quality of the test at the site where it is being performed. While no test is perfect, the knowledge and use of the optimal test in local communities is essential in making the correct clinical decision for your female patients. An understanding of the methods of noninvasive diagnostic testing for CHD and an awareness of how the tests are performed in your community are essential.
Case: Exertional Dyspnea in Elderly, Obese Woman
A 61-year-old, obese woman, height 62" (157.5 cm.), weight 175 lb.(79.5 kg.), was evaluated for exertional dyspnea and chest pain. Her resting ECG showed normal sinus rhythm with non-specific ST-T wave changes. She was able to exercise up to 82% of her maximum potential heart rate (MPHR) before developing chest tightness and shortness of breath. At this point, her stress ECG showed an additional 1.5 to 2.0 mm of ST depression in the inferolateral leads. Tc-99m sestamibi imaging revealed a large perfusion abnormality involving the entire septum, apex, and distal half of the anterior wall. ECG-gated SPECT imaging demonstrated wall motion and thickening which were consistent with severe ischemia involving the left anterior descending (LAD) territory (Fig. 1).
Heart Attacks More Fatal for Women
(see 2 articles below for further information)
.c The Associated Press By LINDA A. JOHNSON
Young women with chest pain should seek immediate medical care, according to researchers who reported today that heart attacks are twice as likely to kill women under age 50 than men in the same age range. The surprising gap narrows and eventually disappears later in life, Yale University researchers found. The discovery suggests that biological factors - not differences in medical care - largely explain why heart attacks are more deadly for women.
Like many studies, this one and a second study in today's New England Journal of Medicine found that women's heart attacks in general are more likely to be fatal. The Yale doctors found 17 percent of female heart attack victims die while still in the hospital, compared with 12 percent of males. However, when the researchers broke the numbers down by age, they came to the surprising conclusion that the difference results entirely from a much higher death rate among the younger victims.
``Heart disease is the No. 1 killer of women, even among younger women, less than 65,'' said Dr. Viola Vaccarino, who directed Yale's study. It was based on a review of the records of 384,878 heart attack victims between 1994 and 1998. Among patients under age 50, when heart attacks are especially rare among women, just 3 percent of the male victims died, compared with 6 percent of females. By age 75, the death rate for both sexes is about equal, at 19 percent.
Over the years, experts have speculated about why heart attacks are more deadly for women, and some have suggested that differences in medical treatment play a role. While this may be part of the story, the new work suggests that biology is probably a more important factor.
``Probably biological mechanisms play a major role, but we need to look at the big picture and take into account all aspects of the women and their care,'' Vaccarino said.
In earlier work, the Yale team and other researchers found that women wait hours longer after a heart attack before going to the hospital, then are treated less aggressively than men. That delay, which allows further damage to the oxygen-starved heart, results partly because women tend to experience less painful heart attack symptoms. Sometimes they feel only pressure or a burning feeling, not crushing pain.
The researchers also noted that younger female victims are more likely than men to have other health problems, such as diabetes, high blood pressure and heart failure.
Still, the treatment differences and women's poorer overall health only explained one-third of the mortality difference. Those factors being equal, women had a 7 percent higher risk of dying for each five years below age 75. By age 80, the men had a slightly higher risk.
In an accompanying editorial, Dr. Laura Wexler of the University of Cincinnati called the discovery ``striking and new.'' ``Do these women have some especially potent risk factor, or do they lack a protective factor that is normally present in women?'' she asked.
Typically heart disease afflicts women about a decade later in life than men, probably because the female hormone estrogen protects their hearts through youth and middle age. Those who are struck at an early age may have differences in the way their bodies use this hormone.
The second study, directed by Dr. Judith Hochman of St. Luke's-Roosevelt Hospital Center in New York, looked at 12,142 men and women who had bad heart attacks, milder ones or severe chest pain. In all three categories, women were up to twice as likely to suffer serious complications. Among those who had heart attacks, the women were 50 percent more likely to die within 30 days.
``Taken together, these studies suggest that there are gender-specific factors at play in heart attacks that are not necessarily related to differences in care but to the underlying disease process,'' said Dr. Lynn Smaha, president of the American Heart Association. Whatever the reasons, several experts stressed the need for women to seek testing and treatment at the first hint of heart trouble and to try to prevent it by controlling their weight, exercising and quitting smoking.
``We really have to look for new treatments'' because the current ones, based mostly on research on men, don't work as well on women, added Dr. Kevin Schulman of Georgetown University Medical Center in Washington.
Heart Bias Study Was Misinterpreted
.c The Associated Press
BOSTON (AP) - The editors of the New England Journal of Medicine say they ``take responsibility'' for media reports which greatly exaggerated conclusions in a study about possible gender and sex bias in heart care.
The study, published in the Journal on Feb. 25, reported what happened when doctors viewed taped interviews of actors describing their identical symptoms and were asked what treatment they would recommend.
It found that, in cases of equally sick patients, doctors were less likely to refer blacks and women than they were whites and men to have cardiac catheterization, a test used to diagnose heart disease.
Several news organizations, including the AP, interpreted the study to show that doctors were 40 percent less likely to order the tests for women and blacks than for men and whites.
However, a followup published in the Journal recently concluded that the likelihood of women and blacks being referred for the tests was actually 7 percent less than for men and whites.
The followup, written by Dr. Lisa M. Schwartz and others from the VA Outcomes Group in White River Junction, Vt., said the misunderstanding resulted from the original study's use of an ``odds ratio'' to report the differences rather than a more commonly used ``risk ratio.''
The researchers calculated the odds in favor of blacks being offered the test and of whites being offered the test. Then they calculated the ratio of these two figures. The ratio of blacks' odds to whites' odds worked out to 0.6, as did the ratio of women's odds to men's.
The media interpreted this to mean that women and blacks were 40 percent less likely to be offered catheterization. But the true difference is much smaller. A table published with the study shows that actually 85 percent of women and blacks were referred for catheterization as were 91 percent of men and whites. This means that the risk ratio was .93. In other words, the probability of referral was 7 percent lower for blacks and women than for whites and men.
The journal editors said they ``take responsibility for the media's overinterpretation'' of the study's findings and said they should not have allowed the use of odds ratios in the study's summary.
AP Corrects Heart Bias Story
.c The Associated Press
BOSTON (AP) - A study's use of an unusual statistical method led The Associated Press to misreport and greatly exaggerate conclusions about possible gender and sex bias in heart care in a Feb. 24 story.
Several news organizations, including the AP, interpreted the study, published in the New England Journal of Medicine, to show that doctors were 40 percent less likely to order cardiac catheterization tests for women and blacks than for men and whites.
In fact, the likelihood of women and blacks being referred for the tests was 7 percent less than for men and whites.
The journal editors said they ``take responsibility for the media's overinterpretation'' of the study's findings and said they should not have allowed the use of misleading statistics which appeared to support the 40 percent figure.
72nd Scientific Sessions of the American Heart Association
Day 2 - November 8, 1999
Stephen J. Lahey, MD, David J. Malenka, MD, William C. Nugent, MD, Christopher Maloney, MD, Bruce Leavitt, MD, Edward R. Nowicki, MD, Robert Clough, MD, Hebe B. Quinton, MS, Daniel J. O'Rourke, MD, Gerald T. O'Connor, PhD, Northern New England Cardiovascular Disease Study Group, Lebanon, New Hampshire
Much has been written about gender as an independent risk factor for surgical mortality risk among patients undergoing coronary artery bypass graft surgery (CABG). The issue remains unresolved; several recent publications offer opposing conclusions. A variety of biologic, demographic, and social factors may influence the short-term, perioperative outcome, but little is known about long-term outcome in women who have survived CABG as compared with men. Most reports document lower operative mortality rates for men and, especially, women, which may be attributable to newer surgical techniques and methods of myocardial protection. However, several questions persist:
Do women receive medical attention later in the course of their disease?
Are women treated less aggressively because of differences in outcome?
Is the current, risk-adjusted, in-hospital mortality comparable for women and men?
Is 5-year survival comparable in women and men?
In this study, we report the 5-year survival for a contemporary, consecutive series of women and men undergoing CABG surgery in Northern New England between 1992 and 1996. By linking the clinical database of the Northern New England Cardiovascular Disease Study Group to the National Death Index, long-term survival after CABG was determined.
The objectives of this study were: (1) to define long-term outcome after CABG in women as compared with men, and (2) to determine whether factors known to adversely affect in-hospital mortality had a similar affect on long-term survival.
Methods
Clinical data were acquired from 14,422 consecutive patients undergoing CABG surgery in our region and entered into the Northern New England Cardiovascular Disease Study Group CABG Registry. Long-term survival was determined by linking this dataset to the National Death Index.
A Student t test was used to compare mean values. A chi-square test was used for comparison of proportions and a Cox proportional hazard regression allowed for comparison of survival between men and women. Long-term survival between men and women was compared by constructing Kaplan-Meier curves.
Results
In this contemporary, consecutive series of patients undergoing CABG, 4021 women and 10,401 men contributed 31,219 person-years of follow-up (women = 8513; men = 22,777). There were 1242 deaths at 5 years (women = 392; men = 850) for a crude, overall, 5-year mortality rate of 4% (4.6% for women and 3.7% for men).
Although the frequency of 3-vessel disease was similar for men and women, several significant differences in observed adverse preoperative risk factors were noted. At the time of surgery, women were older (women, mean age of 67.9 years; men, mean age of 64.1 years), more often had diabetes (35.5% vs 25%), more frequently had peripheral vascular disease (18% vs 15.3%), and more often had a left ventricular end-diastolic pressure greater than 22 mm Hg (29.8% vs 22.9%). Women were also more likely to have congestive heart failure at the time of surgery (17.2% vs 11.1%) and were more likely to require urgent or emergency surgery (69.8% vs 60.8%). Women were, however, less likely to have an ejection fraction of less than 40% (women, 12.7%; men, 15.7%).
An analysis of the crude and adjusted mortality rates for men and women as defined by both the in-hospital mortality and the 30-day mortality yielded interesting findings. Significant differences in the crude in-hospital mortality (men, 2.83%; women, 4.18%; P< .001) and 30-day mortality (men, 4.66%; women, 6.07%; P= .001) are seen between women and men. However, when adjusted for age, peripheral vascular disease, renal failure, preoperative congestive heart failure, preoperative ejection fraction, preoperative left ventricular end diastolic pressure, reoperation, and surgical priority, no difference in the in-hospital and 30-day mortality rates for women and men was observed.
The 5-year crude and adjusted Kaplan-Meier survival curves are similarly revealing. Women have a slightly worse survival curve, with a hazards ratio (women vs men) of 1.22, when crude data are analyzed. However, when survival data are adjusted for the previously cited preoperative comorbidities, no difference is seen in the Kaplan-Meier survival curves.
Discussion
Long-term survival after CABG has been difficult for clinicians to determine. However, by linking data from the National Death Index to our regional database of the Northern New England Cardiovascular Disease Study Group, we have been able to construct accurate Kaplan-Meier survival curves post-CABG in men and women. Several conclusions may be drawn from this large study. First, women present with more concomitant conditions and with more acute conditions than men. Second, although the crude in-hospital and 30-day mortality rates are higher in women, those rates, when adjusted for many of the more common adverse risk factors, are comparable in women and men. Finally, when the 5-year survival curves are adjusted for those same comorbidities, no difference in long-term outcome between women and men is observed. The size of the database suggests that these observations cannot be explained by chance, bias, or confounding variables.
Despite the important implications of this study, several limitations must be kept in mind. This study reports on the regional experience of the Northern New England Consortium and, as such, it is impossible to determine whether these results may be generalized to other regions of the country. Further, we could not determine the impact of intercurrent care during the study period. Nevertheless, we believe that, based on data presented in this report, gender as an independent factor has no appreciable influence on outcome after CABG.
Back to The WAVE~~~