Current diagnosis-related group–based bundling for upper-extremity arthroplasty: a case of insufficient risk adjustment and misaligned incentives

Abstract

The current Centers for Medicare & Medicaid Servicesdiagnosis-related group (DRG) bundled-payment model for upper-extremity arthroplasty does not differentiate between the type of arthroplasty (anatomic total shoulder arthroplasty [ATSA] vs. reverse total shoulder arthroplastyvs. total elbow arthroplasty [TEA] vs. total wrist arthroplasty) or the diagnosis and indication for surgery (fracture vs. degenerative osteoarthritis vs. inflammatory arthritis). The 2011-2014 Medicare 5% Standard Analytical Files (SAF5) database was queried to identify patients undergoing upper-extremity arthroplasty under DRG-483 and -484. Multivariate linear regression modeling was used to assess the marginal cost impact of patient-, procedure-, diagnosis-, and state-level factors on 90-day reimbursements. Of 6101 patients undergoing upper-extremity arthroplasty, 3851 (63.1%) fell under DRG-484 and 2250 (36.9%) were classified under DRG-483. The 90-day risk-adjusted cost of an ATSA for degenerative osteoarthritis was $14,704 ± $655. Patient-level factors associated with higher 90-day reimbursements were male sex (+$777), age 75-79 years (+$740), age 80-84 years (+$1140), and age 85 years or older (+$984). Undergoing a TEA (+$2175) was associated with higher reimbursements, whereas undergoing a shoulder hemiarthroplasty(-$1000) was associated with lower reimbursements. Surgery for a fracture (+$2354) had higher 90-day reimbursements. Malnutrition (+$10,673), alcohol use or dependence (+$6273), Parkinson disease (+$4892), cerebrovascular accident or stroke (+$4637), and hyper-coagulopathy (+$4463) had the highest reimbursements. In general, states in the South and Midwest had lower 90-day reimbursements associated with upper-extremity arthroplasty. Under the DRG-based model piloted by the Centers for Medicare & Medicaid Services, providers and hospitals would be reimbursed the same amount regardless of the type of surgery (ATSA vs. hemiarthroplasty vs. TEA), patient comorbidity burden, and diagnosis and indication for surgery (fracture vs. degenerative pathology), despite each of these factors having different resource utilization and associated reimbursements. Lack of risk adjustment for fracture indications leads to strong financial disincentives within this model.