International Journal of COPD (2019-09-01)

Evaluation of rescue medication use and medication adherence receiving umeclidinium/vilanterol versus tiotropium bromide/olodaterol

  • Moretz C,
  • Bengtson LGS,
  • Sharpsten L,
  • Koep E,
  • Le L,
  • Tong J,
  • Stanford RH,
  • Hahn B,
  • Ray R

Journal volume & issue
Vol. Volume 14
pp. 2047 – 2060


Read online

Chad Moretz1, Lindsay GS Bengtson2, Lucie Sharpsten2, Eleena Koep2, Lisa Le2, Junliang Tong2, Richard H Stanford1, Beth Hahn1, Riju Ray1 1Glaxo Smith Kline, Research Triangle Park, Durham, NC, USA; 2Optum, Eden Prairie, MN, USACorrespondence: Chad MoretzGlaxoSmithKline, Research Triangle Park, 5 Moore Drive, Durham, NC, USATel +1 828 442 6294Email [email protected]: This was the first real-world head-to-head study comparing inhaled long-acting muscarinic antagonist/long-acting β2-agonist fixed-dose combination treatments as maintenance therapy.Methods: Retrospective observational study including commercial, Medicare Advantage with Part D or Part D-only enrollees aged ≥40 years from the Optum Research Database. Patients initiated umeclidinium/vilanterol (UMEC/VI) or tiotropium bromide/olodaterol (TIO/OLO) between June 1, 2015 and November 30, 2016 (index date) with 12 months of pre- and post-index continuous enrollment. Outcomes were modeled following the inverse probability of treatment weighting. The primary endpoint, rescue medication use, was modeled using weighted ordinary least squares regression with bootstrapped variance estimation. Intent-to-treat analysis evaluated non-inferiority and superiority of UMEC/VI to TIO/OLO with thresholds of 0.30 and 0 units, respectively. On-treatment sensitivity analysis evaluated the superiority of UMEC/VI to TIO/OLO for rescue medication use. The secondary endpoint, medication adherence (proportion of days covered [PDC]≥80%), was evaluated using weighted logistic regression. Post hoc weighted Cox proportional hazards regression analysis evaluated escalation to multiple inhaler triple therapy (MITT).Results: The study population included 14,324 patients; 9549 initiated UMEC/VI and 4775 initiated TIO/OLO. During the 12-month post-index period, UMEC/VI initiators used 0.16 fewer adjusted mean units of rescue medication than TIO/OLO initiators (95% CI: −0.28, −0.04), meeting pre-specified non-inferiority (P<0.001) and superiority (P=0.005) criteria; the on-treatment sensitivity analysis for superiority was not statistically significant. Significantly more UMEC/VI than TIO/OLO initiators (28.6% vs 22.7%; P<0.001) achieved a clinically meaningful level (PDC≥80%) of medication adherence. The adjusted risk of escalation to MITT was similar between treatment groups (HR=0.93; 95% CI: 0.81, 1.06; P=0.268).Conclusion: UMEC/VI was superior to TIO/OLO for rescue medication use and UMEC/VI initiators had better medication adherence than TIO/OLO initiators. This study supports findings from a head-to-head trial that demonstrated significant, clinically meaningful improvements in lung function with UMEC/VI versus TIO/OLO.Keywords: COPD, long-acting β2-agonists, long-acting muscarinic antagonists, adherence, rescue medication, real world