Background/Purpose: Injectable biologics are commonly used to treat patients (pts) with moderate to severe rheumatoid arthritis (RA); the frequency with which they are prescribed but not filled is unknown. This exploratory analysis aimed to evaluate filling of newly prescribed injectable biologics for RA and characterize pt outcomes.

Methods: In a retrospective cohort design, pts (age ≥18 years during study period) with an RA diagnosis (ICD9: 714.XX) in 2007-2013 were selected from a de-identified database of clinical information from electronic health records (EHR; Humedica) linked to healthcare claims (Optum) from commercial and Medicare Advantage health plans. First injectable biologic prescription date in EHR was the index date. Pts without continuous pharmacy coverage for ≥6 months pre- and post- index, with evidence of pre-index injectable biologic administration in EHR or claims, or with hospitalization within 30 days post-index were excluded. Pts were categorized as filling the biologic prescription within 30 days (early fillers), 31-180 days (late fillers), or not at all within 180 days (non-fillers) of index. Pt baseline characteristics, including claims-based index of RA severity, RA prescribing patterns, and pt-reported pain scores (0-10; provider-determined scales) from EHR were assessed across all pts; 6-month post-index healthcare resource utilization (including biologic fills) and costs as identified within claims were assessed in pts with continuous medical and pharmacy coverage.

Results: Of 381 pts meeting inclusion criteria, 171 (45%) and 60 (16%) filled an injectable biologic prescription within 30 days and 31-180 days of index, respectively; 90% of prescriptions were written for TNF inhibitors (TNFi). Early fillers were younger, more likely to be female, had higher baseline RA severity, and filled more prescriptions for any reason pre-index. Of non-fillers, 65% were Medicare pts vs 18 and 37% of early and late fillers, respectively. Filling of nonbiologic DMARD prescriptions within 30 days of index was highest in early biologic fillers (45.6%) and lowest among non-fillers (23.3%); however, during days 31-180, the rate was 5.9% in early biologic fillers vs 34.0% in non-fillers. Of early fillers, 14% did not have another biologic prescription filled after 30 days. In the small subgroup of pts with both pre- and post-index pain scores, mean pain scores decreased in early fillers (-1.6; n=11), but increased in late fillers (0.5; n=6) and non-fillers (1.1; n=7). In pts with pharmacy and medical coverage for 180 days post-index (n=375), early fillers had greater RA-related pharmacy (84% of cost difference) and medical resource use and costs than late and non-fillers combined.

Conclusion: Over half (55%) of pts prescribed injectable biologics (90% TNFi) did not fill the prescription within 30 days; 40% had not filled by 180 days. For evaluation of clinical outcomes, documentation of pain scores in the EHR was limited for structured analysis. As expected, healthcare resource use/costs were higher in early fillers over short-term observation. Future research will need to focus on long-term consequences of under-treatment on clinical and economic outcomes.