Abstract 180: Variation in the Use of Angiography Between Procedural and Non-procedural Cardiologists

Background: Physician specialty is associated with adherence to practice guidelines; clinical activities may influence management approaches when guidelines create equipoise. In a scenario in which either a non-invasive stress test or invasive angiography is appropriate, we hypothesized that cardiologists who perform cardiac procedures would be more likely to recommend an invasive approach compared to non-procedural cardiologists. Methods: We developed a web-based, email survey with a 2-part standardized patient case. The first part described a patient with symptoms suggestive of obstructive coronary artery disease (CAD) the second part described an abnormal exercise treadmill test (ETT) result in that patient. In both parts, the scenario was designed to represent an intermediate likelihood of obstructive CAD. Providers were asked to rate the likelihood of obstructive CAD and choose between angiography and stress testing for that patient; both a secondary stress and angiography would be consistent with guidelines. Cardiologists were classified based on self-report to non-procedural (general, non-invasive, heart failure/transplant, congenital) or procedural (invasive, interventional, electrophysiology and cardiothoracic surgery) practice. Multivariable models were used to determine the association between procedural practice and referral for angiography controlling for provider factors (age, gender, years in practice, practice setting) and estimated likelihood of CAD for that patient. Results: Of the 500 cardiologists who responded, 41.8% were procedural cardiologists. Procedural cardiologists similarly rated the patient as having a high likelihood for CAD (part 1: 38.8% vs 40.9%, p= 0.63; part 2: 81.8% vs 81.1%, p = 0.84) compared to non-procedural cardiologists. Although for a patient with intermediate risk of CAD by symptoms, more procedural cardiologists recommended direct angiography referral (12% vs 5.6%, p <0.01) than non-procedural cardiologists. In multivariable models, procedural practice remained associated with higher direct angiography referral (OR 2.67; 95% CI 1.30, 5.49, p= 0.008). After an intermediate risk ETT result, both groups recommended angiography more often (70.5% vs 68.2% p=0.58) than secondary stress testing. Conclusions: Procedural cardiologists more often referred a patient with symptoms concerning for CAD directly for angiography compared to non-procedural cardiologists. When presented an abnormal stress test, both groups were equally likely to recommend angiography over additional stress testing. Therefore, clinical activities may influence management approaches; procedural cardiologists pursued an invasive approach earlier in the diagnostic pathway. Efforts to understand variation in invasive procedure use should take physician clinical activities into account.