Impact of system and physician factors on the detection of obstructive coronary disease with diagnostic angiography in stable ischemic heart disease.

BACKGROUND Wide variation exists in the detection rate of obstructive coronary artery disease (CAD) with elective coronary angiography for suspected stable ischemic heart disease. We sought to understand the incremental impact of nonclinical factors on this variation. METHODS AND RESULTS We included all patients who underwent coronary angiography for possible suspected stable ischemic heart disease, from October 1, 2008, to September 30, 2011, in Ontario, Canada. Nonclinical factors of interest included physician self-referral for angiography, the physician type (invasive or interventional), and hospital type. Hospitals were categorized into diagnostic angiogram only centers, stand-alone percutaneous coronary intervention centers, or full service centers with coronary artery bypass surgery available. Multivariable hierarchical logistic models were developed to identify system and physician-level predictors of obstructive CAD, after adjustment for patient factors. Our cohort consisted of 60 986 patients, of whom 31 726 had obstructive CAD (52.0%), with significant range across hospitals from 37.3% to 69.2%. Fewer self-referral patients (49.8%) had obstructive CAD compared with nonself-referral patients (53.5%), with an odds ratio of 0.89 (95% confidence interval, 0.86-0.93; P<0.001). Angiograms performed by invasive physicians had a lower likelihood of obstructive CAD compared with those by interventional physicians (48.2% versus 56.9%; odds ratio, 0.85; 95% confidence interval, 0.81-0.90; P<0.001). Fewer angiograms at diagnostic only centers showed obstructive CAD (42.0%) compared with full service centers (55.1%; odds ratio, 0.62; 95% confidence interval, 0.39-0.98; P=0.04). Nonclinical factors accounted for 23.8% of the variation between hospitals. CONCLUSIONS Physician and system factors are important predictors of obstructive CAD with coronary angiography.