Background/Purpose: RA-related hospitalization and surgery (e.g. total joint arthroplasty [TJA]) are important long-term outcomes in RA. While advances in treatment for RA have resulted in decreased admissions for active disease, the national rates of RA-related procedures, hospitalizations and their costs in elderly patients with RA have not been reported. We evaluated the rates of first and subsequent RA-related surgery, hospitalization and mortality in a US Medicare population, analyzed predictors of RA-related surgery and compared total costs between RA patients with and without surgery.

Methods: Data from patients enrolled in the US Medicare Claims Database from January 1999 to December 2009, who had ≥2 RA diagnoses (ICD 714.0) ≥2 months apart during the identification period were analyzed. The date of the first RA diagnosis was designated as the index date. Those with a baseline period <12 months (i.e. those diagnosed during 1999 and 2000) were considered prevalent cases. Kaplan–Meier analysis was used to estimate the cumulative incidence of orthopedic surgery following RA diagnosis. Cox proportional hazards (CPH) modeling identified factors associated with surgery (TJA, TJA-associated procedures, non-TJA) and overall mortality. Estimated healthcare costs (mean and standard deviation) for all RA patients with surgery were compared with those for RA patients without surgery.

Results: The study population comprised 360,912 patients with RA enrolled in Medicare who met the study inclusion criteria. Cumulative 4- and 10-year TJA incidence rates were 7.5 and 13.2%, respectively; mortality rates were 13.2 and 27.9%. Of all RA patients with surgical experience, 86.5% had a TJA procedure. Patient characteristics are shown (Table). Based on CPH models, predictors of surgery varied by type of procedure; positive factors often included regional demographic and co-morbid osteoarthritis (OA) at baseline, and negative predictors often included follow-up therapy and minority race. For example, for TJA, patients with OA and patients living in the Midwest had a higher hazard (hazard ratio [HR]=2.11, p<0.01; HR=1.38, p<0.01, respectively). Patients receiving combination (MTX + biologic DMARD) RA therapies during follow-up had a lower hazard (HR=0.41, p<0.01). RA patients with surgery had almost double the average inpatient cost at $19,382 vs $10,282 for patients without surgery (p<0.01). RA patients with surgery also had significantly higher outpatient, outpatient emergency room, and office costs (p<0.01).

  Conclusion: Medicare patients with RA continue to experience high rates of RA-related surgery, hospitalization and overall mortality. However, biologic DMARD therapies appear to have a protective effect on all outcomes. Greater availability of efficacious RA therapies has the potential to reduce RA-related surgery, hospitalization and costs.