Current issues of ACP Journal Club are published in Annals of Internal Medicine


Review: Exercise tests to detect CAD in women have moderate sensitivities and specificities

ACP J Club. 1999 July-Aug;131:21. doi:10.7326/ACPJC-1999-131-1-021

Source Citation

Kwok Y, Kim C, Grady D, Segal M, Redberg R. Meta-analysis of exercise testing to detect coronary artery disease in women. Am J Cardiol. 1999 Mar 1;83:660-6.



What is the accuracy of exercise electrocardiography (ECG), exercise thallium scanning, and exercise echocardiography compared with angiography for detecting coronary artery disease (CAD) in women?

Data sources

2 MEDLINE searches were done; both used the terms coronary disease diagnosis and exercise test. The first (1966 to 1995) used the terms exercise electrocardiography, angiography, and heart catherization. The second (1975 to 1995) used terms for exercise echocardiography and exercise radionuclide scanning: thallium or technetium. Bibliographies of relevant reviews and studies were scanned, and experts were contacted.

Study selection

English-language studies were selected if ≥ 50 women were studied and the results of at least 1 exercise test were compared with the results of coronary angiography. Studies that evaluated women after myocardial infarction or angioplasty were excluded.

Data extraction

Data were extracted in duplicate for year of publication; study quality; test characteristics, including exercise protocol; patient numbers and characteristics; test evaluation and angiographic definition of CAD; and percentage of women with confirmed CAD.

Main results

21 studies met the inclusion criteria: 19 evaluated exercise ECG (3721 women), 5 evaluated exercise thallium scanning (842 women), and 3 evaluated exercise echocardiography (296 women). Test characteristics are listed in the Table. All exercise tests had moderate sensitivities and specificities.


Exercise testing methods (electrocardiography, thallium scanning, and echocardiography) have moderate sensitivities and specificities for detecting coronary artery disease in women.

Source of funding: National Institutes of Health.

For correspondence: Dr. R. Redberg, Box 2014, 505 Parnassus Avenue, M1176, School of Medicine, Division of Cardiology, San Francisco, CA 94143-0214, USA. FAX 415-476-0424.

Table. Test characteristics (weighted means) for detecting coronary artery disease in women*

Test Sensitivity (95% CI) Specificity (CI) +LR (CI) -LR (CI)
All studies 61% (54 to 68) 70% (64 to 75) 2.3 (1.8 to 2.7) 0.55 (0.47 to 0.62)
MVD 69% (45 to 94) 61% (48 to 74) 1.8 (1.6 to 1.9) 0.46 (0.17 to 0.77)
End point reached† 66% (60 to 73) 72% (64 to 79) 2.7 (2.1 to 3.3) 0.46 (0.40 to 0.53)
Thallium scanning
All studies 78% (72 to 83) 64% (51 to 77) 2.9 (1.0 to 5.0) 0.36 (0.27 to 0.44)
MVD 85% (80 to 89) 52% (35 to 68) 2.1 (1.0 to 3.9) 0.32 (0.19 to 0.45)
CT 78% (69 to 87) 58% (51 to 66) 1.9 (1.6 to 2.2) 0.38 (0.25 to 0.50)
Planar images 76% (73 to 80) 89% (86 to 92) 6.9 (5.7 to 9.0) 0.26 (0.23 to 0.29)
All studies 86% (75 to 96) 79% (72 to 86) 4.3 (2.9 to 5.7) 0.18 (0.05 to 0.31)

*CT = computed tomography; MVD = multivessel disease. LRs defined in Glossary.
†Patients reached electrocardiographic or heart rate end point.


Because the literature on exercise testing for detection of CAD is very extensive, this meta-analysis is helpful to physicians who use these tests to diagnose CAD in women. Kwok and colleagues use an appropriate study design, and their inclusion criteria and analysis allow us to appreciate the limitations of stress testing in women. A more recent study that evaluated exercise echocardiography reached similar conclusions (1).

Although the 3 exercise tests evaluated in this review are not highly specific or sensitive for detection of CAD in women, a negative test result will continue to have important reassurance value, particularly in the prediction of future clinical cardiac events (2). New developments in stress test methods, such as adding right chest leads, may improve the diagnostic accuracy of the exercise ECG in both sexes (3). When detection of multivessel CAD is emphasized, the sensitivity improves to a more clinically useful level (Table). Although the precise reasons for reduced accuracy of stress testing in women is not known, it is likely that differences in autonomic function, hormonal milieu, function of small coronary vessels, and heart size may play a role. Stress testing should not be abandoned in women, but the results need to be carefully interpreted.

Gary W. Burggraf, MD
Kingston General HospitalKingston, Ontario, Canada


1. Roger VL, Pellikka PA, Bell MR, et al. Sex and test verification bias. Impact on the diagnostic value of exercise echocardiography. Circulation. 1997;95:405-10.

2. Olmos LI, Dakik H, Gordon R, et al. Long-term prognostic value of exercise echocardiography compared with exercise 201TI, ECG, and clinical variables in patients evaluated for coronary artery disease. Circulation. 1998;98:2679-86.

3. Michaelides AP, Psomadaki ZD, Dilaveris PE, et al. Improved detection of coronary artery disease by exercise electrocardiography with the use of right precordial leads. N Engl J Med. 1999;340:340-5.