Cardiovascular disease is the leading cause of mortality for women in the United States. According to the American Heart Association (AHA), approximately one in three female adults have some form of cardiovascular disease. AHA suggests there is evidence showing that women at risk for coronary artery disease (CAD) are less often referred for the appropriate diagnostic test than are men. Coronary anatomy and pathology have traditionally been defined and identified by coronary angiography, a procedure that is indicated in patients who have chest pain and are at high risk for CAD. For intermediate-risk patients, clinicians have a wide range of noninvasive technologies (NITs) to choose from that can assess functional status (i.e., ischemia or no ischemia) or visualize anatomic abnormalities (i.e., no CAD, nonobstructive CAD, or obstructive CAD). Functional modalities include stress electrocardiography (ECG); stress echocardiography (ECHO); and stress radionuclide myocardial perfusion imaging, including single-proton emission computed tomography (SPECT) and positron emission tomography (PET). Anatomic modalities include stress myocardial perfusion and wall-motion cardiac magnetic resonance (CMR) imaging and coronary computed tomography angiography (coronary CTA). The comparative safety and accuracy of these NITs in women was uncertain, although substantial data exists for populations combining men and women, and for mixed populations of known and no known CAD.
In 2012, a Comparative Effectiveness Review (CER), “Noninvasive Technologies for the Diagnosis of Coronary Artery Disease in Women,” evaluated the diagnostic accuracy and risks of NITs in women with symptoms suspicious for CAD, including assessing predictors affecting test accuracy, and the ability of NITs to provide risk stratification and prognostic information, inform decisionmaking about treatment options, and affect clinical outcomes.
A total of 104 comparative studies (110 articles) were included. For women with no known CAD, the summary of accuracy for each NIT modality compared with coronary angiography was ECG (29 studies): sensitivity 62 percent, specificity 68 percent; ECHO (14 studies): sensitivity 79 percent, specificity 83 percent; SPECT (14 studies): sensitivity 81 percent, specificity 78 percent; CMR (5 studies): sensitivity 72 percent, specificity 84 percent; and CTA (5 studies): sensitivity 94 percent, specificity 87 percent. Compared with men evaluated in the same studies, in women ECG and coronary CTA modalities were both less sensitive and less specific. The ECHO and SPECT modalities, although less sensitive, appeared to be more specific in women. The lower specificity of the ECG modality in women was the only statistically significant difference. Strength of evidence was high for ECG, ECHO, and SPECT, and was low for CMR and coronary CTA compared with coronary angiography in women. Eleven comparative studies examined predictors of diagnostic accuracy in women such as postmenopausal status, race/ethnicity, heart size, beta blocker use, and pretest probability; insufficient evidence was available to draw conclusions about predictors that affect accuracy. Eight studies assessed risk stratification and prognostic factors, two studies assessed treatment decisionmaking, and four studies provided comparative clinical outcomes but provided insufficient evidence on the comparative effectiveness of NITs to provide risk stratification, prognostic information, treatment decisionmaking, or impact on clinical outcomes in women. Thirteen comparative studies reported risks. Of these, four studies of coronary CTA showed a higher mean effective radiation dose and attributable risk of cancer incidence in women compared with men; however, radiation safety issues were not discussed in other NIT modalities with radiation exposure. Thus, there was insufficient evidence regarding the comparative risks of various NIT modalities in women.
Given the clinical and economic importance of noninvasive testing for CAD in women, the ongoing investment in NIT research, and the remaining areas of uncertainty, we sought to create a prioritized research agenda that would represent the interests of diverse stakeholders and allow the remaining areas of uncertainty to be addressed.