High probability of disease in angina pectoris patients: is clinical estimation reliable?

According to most current guidelines, stable angina pectoris patients with a high probability of having coronary artery disease can be reliably identified clinically.

To examine the reliability of clinical evaluation with or without an at-rest electrocardiogram (ECG) in patients with a high probability of coronary artery disease.

A prospective series of 357 patients referred for coronary angiography (CA) for suspected stable angina pectoris were examined by a trained physician who judged their type of pain and Canadian Cardiovascular Society grade of pain. Pretest likelihood of disease was estimated, and all patients underwent myocardial perfusion scintigraphy (MPS) followed by CA an average of 78 days later. For analysis, the investigators focused on the approximate groups of patients with more severe disease, ie, typical angina (n=187), Canadian Cardiovascular Society grade 2 pain or higher (n=176) or high (higher than 85%) estimated pretest likelihood of disease (n=142).

In the three groups, 34% to 39% of male patients and 65% to 69% of female patients had normal MPS, while 37% to 38% and 60% to 71%, respectively, had insignificant findings on CA. Of the patients who had also an abnormal at-rest ECG, 14% to 21% of men and 42% to 57% of women had normal MPS. Sex-related differences were statistically significant.

Clinical prediction appears to be unreliable. Addition of at-rest ECG data results in some improvement, particularly in male patients, but it makes the high probability groups so small that the addition appears to be of limited clinical relevance.