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


Using right and left precordial leads increased the sensitivity of exercise testing for detecting coronary artery disease

ACP J Club. 1999 Sept-Oct;131:46. doi:10.7326/ACPJC-1999-131-2-046

Source Citation

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 Feb 4;340:340-5.



Does the addition of the right precordial leads V3R, V4R, and V5R to the standard 12 leads improve the sensitivity of exercise electrocardiography (ECG) for detecting coronary artery disease (CAD)?


A blinded comparison of standard 12-lead exercise ECG (with and without the 3 right precordial leads) with thallium-201 scintigraphy and coronary arteriography.


University hospital in Athens, Greece.


245 patients (mean age 52 y, 89% men) with suspected angina. Exclusion criteria were left or right bundle-branch block, left or right ventricular hypertrophy, ventricular pre-excitation, history of myocardial infarction, valvular or congenital heart disease, previous aortocoronary bypass surgery or coronary angioplasty, or current digitalis treatment.

Description of tests and diagnostic standard

All patients exercised on Quinton 5000 treadmill systems according to the multistage Bruce protocol. Patients had an ECG recorded with the standard 12 leads and an additional 3 right precordial leads (V3R, V4R, and V5R); the results for each set of leads were recorded and analyzed separately. Tomographic thallium 201 scintigraphy was done in all patients. The diagnostic standard was coronary arteriography.

Main outcome measure

Sensitivity and specificity for detecting angiographically confirmed CAD.

Main results

The addition of the 3 right precordial leads had greater sensitivity than did the standard 12-lead technique for detecting all, 1-vessel, and 2-vessel CAD (P < 0.001 for all comparisons) and had sensitivity and specificity similar to those of thallium scintigraphy for detecting all CAD (Table).


The addition of 3 right precordial leads to the standard 12 leads increased the sensitivity of exercise electrocardiography for detecting coronary artery disease.

Source of funding: Not stated.

For correspondence: Dr. A.P. Michaelides, Department of Cardiology, University of Athens Medical School, Hippokration Hospital, Athens, Greece.

Table. Test characteristics for detecting coronary artery disease (CAD)*

Test Sensitivity (95% CI) Specificity (CI) +LR -LR
Right and left precordial ECG leads 92% (88 to 96) 88% (73 to 97) 7.9 0.09
Standard 12 ECG leads 66% (59 to 72) 88% (73 to 97) 5.6 0.39
Thallium scintigraphy 93% (89 to 96) 82% (65 to 93) 5.3 0.09
Type of CAD Sensitivity (CI)
Standard ECG leads Standard and right ECG leads
1-vessel CAD 52% (41 to 63) 89% (81 to 95)
2-vessel CAD 71% (61 to 81) 94% (87 to 98)

*ECG = electrocardiography. LRs defined in Glossary; LRs and CIs calculated from data in article.


Nonimaging exercise tolerance tests are frequently used to determine whether a patient has CAD. The reasonably good specificity of these tests means that few false-positive results occur and that they are usually adequate to rule in disease. Of course, what clinicians really want to know is not whether the test result is positive but what the probability of disease is given a positive result (the positive predictive value). If the pretest probability of disease is low, even a positive result on a nonimaging stress test will not convince clinicians that CAD is present; they will therefore perform additional testing. This is what happens in patients who have atypical angina, no risk factors for CAD, and a positive result on a nonimaging study who then undergo an imaging test.

With only fair sensitivity, nonimaging stress tests are not that effective at ruling out disease. Although they may place the patient in a good prognostic category, patients and clinicians worry about missing the diagnosis of CAD. Michaelides and colleagues' study suggests that by making a minor modification in how the ECG is recorded, the sensitivity of nonimaging exercise testing can be improved to match that of imaging tests; thus, the negative predictive value of the nonimaging tests improves and additional testing may be avoided. One caveat is that the spectrum of disease in this study is likely to be somewhat more severe than that seen in general practice. Therefore, the reported sensitivity will probably be somewhat better than can be expected in primary care settings. Physicians who practice in referral centers will benefit from adopting this new method. Those who work with a less highly referred patient population should not change their practice until additional evidence is available.

David J. Malenka, MD
Dartmouth-Hitchcock Medical CenterLebanon, New Hampshire, USA