The Best Approach to Women with Chest Pain

abstract & commentary

Synopsis: In women with stable chest pain consistent with angina pectoris, stress myocardial perfusion imaging followed by coronary angiography in selected patients is more cost effective in the near term than a cardiac catheterization first strategy.

Source: Shaw LJ, et al. J Nucl Cardiol 1999;6:559-569.

Because of perceived diagnostic inaccuracies with various forms of stress testing for coronary artery disease (CAD) in women, there has been interest in the direct coronary angiography approach. However, this approach has high initial costs. Thus, Shaw and colleagues undertook a multicentered study of 4638 women being evaluated for stable chest pain consistent with angina pectoris. Of these women, 1263 were referred for noninvasive testing first and 3375 for coronary angiography first. Noninvasive testing was exercise or dipyridamole stress SPECT myocardial perfusion imaging. The baseline clinical characteristics, including pretest likelihood of CAD, were not statistically different between the two approaches despite the absence of randomization. The incidence of at least one vessel with greater than 70% diameter stenosis in the angiography first group was 13% of low likelihood patients, 29% of intermediate, and 52% of high likelihood of CAD patients. The incidence of at least one myocardial perfusion defect in the stress test first group was 23% in low-likelihood patients, 27% in intermediate, and 34% in high-likelihood patients. When those with a positive perfusion underwent catheterization, 50%, 55%, and 76%, respectively, had significant coronary lesions. Follow-up for an average of 2.5 years (minimum 6 months from testing) showed death rates from 0.5% to 2.2%, depending on pretest risk, and did not differ between the two approaches. The estimated direct costs per patient for diagnostic testing and follow-up medical care over 2.5 years ranged from $2490 (low likelihood) to $3687 (high likelihood) in the angiography first group and from $1587 to $2585 for the stress testing first group (P < 0.01). Thus, Shaw et al conclude that in women with stable chest pain consistent with angina pectoris, stress myocardial perfusion imaging followed by coronary angiography in selected patients is more cost effective in the near term than a cardiac catheterization first strategy, regardless of pretest likelihood of disease.

Comment by Michael H. Crawford, MD

This observational study suggests that the cost of a noninvasive approach is 30% less than an invasive approach to evaluating women with stable chest pain syndromes consistent with angina, regardless of the clinically estimated likelihood of disease, and that there was no difference in outcome over an average of 2.5 years with either approach. The reasons for this difference include the fact that the invasive approach yielded two times the number of normal coronary angiogram patients as compared to the stress testing first group and the invasive group was more likely to undergo revascularization. Shaw et al argue that this study is strong evidence to pursue stress testing first in women with stable chest pain syndromes.

There are several limitations to this study. The most obvious limitation is the lack of randomization to the two testing strategies. As expected, almost two-thirds of the women were referred for catheterization first, based on current perceptions about the adequacy of noninvasive stress testing in women. Initial clinical characteristics between the two groups showed more high-risk patients and fewer intermediate and low-risk patients in the catheterization first group, although these differences were not statistically significant. One could argue that since outcomes (death and MI) were almost identical between the two groups, they must have been well matched. But remember that more of the catheterization first patients underwent revascularization, which may have improved their outcome. Also, patients with known or almost certain coronary disease were included, as evidenced by fixed perfusion defects in about one-quarter of the patients. Including such patients is known to improve the results of perfusion imaging for predicting who has significant coronary lesions.

The study provides information of interest regarding the perception that stress testing is not particularly accurate in women. Shaw et al acknowledge this perception, but claim that technological advances such as the use of sestimibi in 80% of their patients have largely eliminated this problem. Other technical issues such as prone imaging (avoids breast attenuation) are not discussed. However, the results of stress perfusion imaging in this study are sobering. Of those with a positive study referred for angiography, 33% had normal coronary arteries. A one-third false-positive rate is not reassuring and when stratified for pretest likelihood of disease, specificity ranged from 50-76%. It does not appear that technological improvements in the tertiary care centers participating in this study have improved the false-positive problem in women.

The major difference in the costs are the initial diagnostic testing costs in all risk groups and the higher cost of follow-up care in the high-risk patients undergoing catheterization because of the higher use of revascularization in these patients. However, these are direct costs only. We don’t know about indirect costs such as time lost from work, etc. Also, there was no analysis of quality of life with the two strategies. It is conceivable that quality-adjusted life-years would be better with an invasive approach. In addition, this was a costly, noninvasive approach with nuclear imaging performed in all patients, including low-risk ones. The AHCPR guidelines would not even stress most such patients and many physicians employ ECG stress tests in low-risk women with normal resting ECGs. It is hard to fault the latter strategy since the false-positive rate with ECG couldn’t be much worse than the 50% incidence observed with perfusion imaging in low-risk women in this study. Thus, I remain to be convinced that stress nuclear perfusion imaging for all women with anginal syndromes is the best approach. I’m still going with physician judgment in each individual.

Which of the following is most correct concerning a stress test first vs. a catheterization first approach in women with chest pain consistent with angina?

a. It is clearly superior in high-risk patients.

b. Direct costs are less.

c. The absolute number of normal coronary angiograms is more.

d. Outcomes are superior.