Hospital-Physician Integration and the Comprehensive Care for Joint Replacement Program.

We examined how hospital-physician integration influences hip and knee osteoarthritis treatment under different payment models. We leveraged a natural experiment from Medicare's random assignment of metropolitan statistical areas (MSAs) to the Comprehensive Care for Joint Replacement (CJR) program. In the 12 months after diagnosis, CJR was associated with more primary care physician and orthopedist visits (0.14 and 0.23, p < .001 for both), while integration was associated with fewer (-0.16 and -0.59, p < .001 for both). We discovered an extensive margin effect: CJR was associated with lower likelihood of joint replacements (-1.9 percentage points [pp], p < .001); integration was associated with higher likelihood (1.0 pp, p < .001). Among knee replacement patients, those in CJR MSAs were less likely to receive surgery in outpatient settings (-4.6 pp, p < .001). Payment models and organizational structure may create opposing incentives, where bundled payment decreases and vertical integration increases the likelihood of joint replacement surgery.