How well does standard clinician evaluation identify low likelihood of ischaemic or coronary heart disease?

Guidelines claim that patients with a low likelihood of coronary artery disease can be reliably identified clinically by a simple set of rules. Among 385 patients referred for coronary angiography because of suspected stable angina pectoris we found by myocardial perfusion scintigraphy in three selected low likelihood groups reversible perfusion defects in 23%-29% of male and 11%-17% of female patients. Rates of significant angiographic disease were similar. These data question the validity of current clinical practice for identifying low likelihood of disease in this category of patients.