Background/Purpose: In 2015, the Center for Medicare and Medicaid Services released claims data on payments to over 950,000 health care providers paid through Medicare Part B. This study investigated 2013 Part B utilization and costs among rheumatology providers in the state of Illinois, and links between utilization data and local demographics.

Methods: Publicly available 2013 Medicare Part B reimbursement data was compiled through ProPublica Treatment Tracker (https://projects.propublica.org/treatment/). Providers identified by Medicare as rheumatology providers in Illinois (n=191) with individual practitioner payment and service data were used for analysis. Practice location as defined by Medicare-assigned zip code was used to search 2013 census data ( http://www.census.gov) for geographic distributions of race/ethnicity, education, and income. Six provider level outcomes were investigated. Both services delivered and payments were examined for: A) rheumatoid arthritis (RA) biologics; B) osteoarthritis (OA) viscosupplements, and; C) totals billed to Medicare. T-tests investigated differences in mean outcomes between providers who did versus did not order services. Multivariable regression analyses investigated geographic factors associated with any ordering (0/1; logistic regressions) and continuous payments (generalized linear models; gamma family, log link) of RA biologics and viscosupplements.

Results: Total of 191 Illinois rheumatology providers were identified by Medicare for 2013. The top ten providers accounted for 26.5% of total Medicare payments and 32% total number of services. Significant differences between Part B RA biologic utilizers (n=38) and non-utilizers (n=153) existed for: total number of services (mean 42763 vs 4598, p<0.001); services per patient (81 vs 15, p<0.001); total payments ($770,342 vs 108,196, p<0.001); and payment per patient ($1,510 vs 333, p<0.001). Similar data existed for viscosupplements utilizers (n=44) vs non-utilizers (n=147) in terms of: total number of services (mean 3213 vs 6257, p<0.001); services per patient (54 vs 20, p<0.001); total payments ($600,052 vs 132,141, p<0.001); and payment per patient ($1,048 vs 423, p<0.001). There were correlations between biologic utilization for RA and viscosupplements for OA in both total payments (rho = .52, p<0.001) as well as number of services (rho = .48, p<0.001). In adjusted regression analyses, a 1% zip code-level increase in percent white race/ethnicity was associated with both a 4% increase in the odds of any RA biologics orders (p=0.002) and a 6% increase in RA biologics payments (p<0.001).

Conclusion: We observed dramatic differences in 2013 Medicare Part B utilization among rheumatology providers in Illinois for RA biologic and OA viscosupplement use. In adjusted regressions analysis, percent white race was positively associated with RA biologic use. Further research is needed to better understand health care utilization in rheumatology and its relation to population demographics. This study did not investigate quality of care or patient outcomes and did not examine medications not covered under Part B. Payments do not reflect provider profit as providers may incur substantial overhead costs.

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/medicare-part-b-utilization-by-rheumatology-health-care-providers-as-related-to-population-demographics-from-2013/