Analysis of physician and hospital differences in 'negative' coronary angiogram rates.

A direct relationship has been postulated between high "negative" coronary angiogram rates and physician payment. We conducted a prospective study of coronary angiography in a teaching and community hospital staffed, respectively, by cardiologists who were performing cardiac catheterization as salaried or fee-for-service physicians. The lower overall rate of negative angiograms at the teaching hospital correlated with the presence of a cardiac surgery unit and the increased referral of patients with documented coronary artery disease. The percentage of completely normal angiograms did not differ significantly between hospitals. The number of angiograms positive by a 70% occlusion criterion in patients not previously known to have coronary artery disease also did not differ greatly. Negative angiogram rates appeared to vary inversely with physician ability to set preangiogram probabilities of coronary artery disease. Our findings do not discount reimbursement as a strong incentive, but suggest other important determinants of coronary angiographic variation.