Objective: To determine factors contributing to the relatively high frequency and variability (10% to 30%) of finding no significant coronary disease by coronary angiography in patients with chest pain.
Study design: Retrospective, comparative analysis of practice patterns at 3 southeastern Michigan hospitals and a composite sample from New York State.
Patients and methods: Medical records for 7668 patients were reviewed to determine the frequency of negative coronary arteriographic findings in patients undergoing chest pain evaluation. A private practice allopathic community hospital with interventional cardiologists and a private practice osteopathic community hospital with diagnostic facilities (DiagCommunity) were compared with a university hospital with full-time salaried interventional cardiologists and a sample of 17 New York hospitals.
Results: Of the 7668 coronary angiograms at all centers, 39.7% were performed to assess patients with stable chest pain. There was no significant obstruction found in 16.5%, and the frequency was not different between the Michigan (17.8%+/-3.8%) and New York (14.2%) hospitals. The DiagCommunity had the highest proportion (22%; P < .001 vs others). On review of the negative coronary arteriographic findings, normal or near normal coronary arteriographic findings were infrequent (range, 2.4%-6.6%) but higher in the DiagCommunity (6.6% vs 2.9%+/-1.6%; P < .0001).
Conclusions: The frequency of finding no significant coronary disease by arteriography in patients with chest pain is similar in southeastern Michigan hospitals and comparable to an established external database. Cardiology self-referral and personal gain does not seem to be a major factor in selection of patients for invasive studies.