Background/Purpose: Variations in biologic prescribing habits by rheumatology providers may account in part for variability in direct cost of care for patients with rheumatoid arthritis (RA).  We used data from the Rheumatology Informatics System for Effectiveness (RISE) registry to estimate variability in biologic use among US rheumatologists.

Methods: RISE is a national, EHR-enabled registry that passively collects and houses data on all patients seen by participating practices and thereby avoids sampling bias.  As of December 2016, RISE was connected to 632 providers (99.5% of which were physicians) representing an estimated 16% of the US workforce.  We calculated, on a per-provider basis, the proportion of RA patients prescribed a biologic or tofacitinib at least once between January and December 2016 (inclusive).  The population of patients in the denominator, numbering 58,055 across all participating practices, was defined as those assigned an ICD code for RA in at least two separate encounters in 2016.  Consistent with other national performance analyses, providers who saw fewer than 30 RA patients (315 of the 632 RISE-connected providers) were excluded.  Few patients saw >1 provider, but in those instances prescriptions were attributed only to the provider with the plurality of RA-coded visits; therefore, each patient was attributed to only one provider.

Results: Provider-to-provider variability in the proportion of RA patients who received a biologic prescription is shown in Figure 1.  Across the US in 2016, an average of 38% of each provider’s RA patients were prescribed a biologic but the fraction of patients prescribed a biologic ranged from 0 to 79%.  A regional breakdown is shown in Table 1 and demonstrates statistically significant geographic variability as well.

Conclusion: These data estimate the degree to which biologic prescription patterns vary across the US.  In light of existing data indicating that biologics account for the majority of the direct cost in providing care for US patients with RA, and ongoing initiatives such as the Merit Incentive Payment System to incorporate measures of resource utilization in payment reform, this study provides initial benchmarking information for rheumatology providers in the US.  Our study does not address quality of care.  Future studies adjusted for case mix and disease activity will be required to estimate variability in the value (quality divided by cost) of care.

Figure 1.  Fraction of RA patients prescribed biologics, by provider.  Red line indicates the mean.

Table 1.  Fraction of RA patients prescribed biologics, on a per-provider basis, by region.