To examine the influence of physician and hospital market structures on medical technology diffusion, studying the diffusion of drug‐eluting stents (s), which became available in April 2003.
Data Sources/Study Setting
Medicare claims linked to physician demographic data from the American Medical Association and to hospital characteristics from the American Hospital Association Survey.
Retrospective claims data analyses.
Data Collection/Extraction Methods
All fee‐for‐service Medicare beneficiaries who received a percutaneous coronary intervention () with a cardiac stent in 2003 or 2004. Each record was joined to characteristics on the patient, the procedure, the cardiologist, and the hospital where the was delivered. We accounted for the endogeneity of physician and hospital market structure using exogenous variation in the distances between patient, physician, and hospital locations. We estimated multivariate linear probability models that related the use of a in the on market structure while controlling for patient, physician, and hospital characteristics.
DESs diffused faster in markets where cardiology practices faced more competition. Conversely, we found no evidence that the structure of the hospital market mattered.
Competitive pressure to maintain or expand volume shares compelled cardiologists to adopt s more quickly.
