Noninvasive diagnosis of coronary artery disease in women has some limitations due to a higher percentage of false-positive results. In addition, a lower prevalence of disease can be observed in this population. In this study, we evaluated the diagnostic accuracy of 99mTc-methoxy-isobutyl-isonitrile (MIBI) SPECT in women and men, in a group of patients with proven coronary artery disease by coronary angiography (select minority) and in all patients where a noninvasive test (silent majority) was performed.
Methods: Seven hundred and two consecutive patients without previous myocardial infarction were studied with 99mTc-MIBI myocardial SPECT. One hundred sixty-three patients had coronary angiography (select minority) and 539 did not (silent majority). All patients underwent exercise stress testing, and simultaneous dipyridamole was administered in 32% of patients who did not achieve maximum predicted heart rates. Diagnostic accuracy of the test was calculated for the select minority. Then, sensitivity and specificity of the silent majority were recalculated according to the Diamond criteria.
Results: Prevalence of coronary artery disease (32% versus 80%, p = 0.0001) and peak O2 consumption achieved in exercise tests (watts, exercise duration) were lower in women. The probability of positive results of 99mTc-MIBI SPECT also was lower in women (34% versus 65%). The sensitivity of 99mTc-MIBI SPECT in women of the select minority was lower (85% versus 93%, p = 0.01), whereas there was no significant difference for specificity (91% versus 89%). After correcting the results for the silent majority, there were no significant differences in sensitivity (87% versus 88%) and specificity (91% versus 96%) between women and men. These results were not different for patients who achieved maximum predicted heart rates during stress testing (without dipyridamole administration).
Conclusion: The sensitivity of 99mTc-MIBI myocardial SPECT in women was lower than in men when only the select minority was considered. When the silent majority was considered (correction of selection bias) sensitivity and specificity results did not differ significantly between the sexes.