Real-World Comparative Effectiveness Study in Patients with Asthma Initiating Fluticasone Furoate/Vilanterol or...

We compared the real-world effectiveness of initiating beclometasone dipropionate/formoterol fumarate (BDP/FOR) versus fluticasone furoate/vilanterol (FF/VI) in a general practice (GP) asthma cohort in England.

Patients newly initiating BDP/FOR or FF/VI between 1 December 2015 and 28 February 2019 (index), were selected from anonymised Clinical Practice Research Datalink data. Baseline was < 12 months pre-index with ≤ 12 months follow-up post-index. Eligible patients were aged ≥ 18 years at index, had diagnosed asthma, ≥ 1 FF/VI or BDP/FOR prescription, medical records eligible for linkage to secondary care data and continuous GP-registration ≥ 12 months pre-index. Patients with chronic obstructive pulmonary disease, ≥ 1 fixed-dose inhaled corticosteroid/long-acting β₂-agonist, single-inhaler triple or biologic therapy at index were excluded. The primary study outcome was asthma exacerbation rate. Secondary outcomes included medication persistence and oral corticosteroid (OCS) use. Propensity scores were generated for each treatment comparison; inverse probability of treatment weighting adjusted for confounding in baseline characteristics between groups, applied to each outcome separately. Analyses considered intercurrent events (ICEs; treatment switching, discontinuation, loss to follow-up, death, rescue medication use).

Weighted group standard mean differences showed adequate balance for most covariates. Patients initiating BDP/FOR (n = 46,809) and FF/VI (n = 3773) had numerically similar exacerbation rates per person per year (PPPY) while-on index treatment [measuring outcome until ICE; BDP/FOR, 0.1479 (n = 31,715); FF/VI, 0.1338 (n = 2547); rate ratio 0.9048, p = 0.2841]. Patients continuing uninterrupted index treatment for 12 months had a lower exacerbation rate PPPY for FF/VI [0.0681 (n = 384)] than BDP/FOR [0.1104 (n = 3342); rate ratio, 0.6162 (p = 0.0293)]. For patients initiating FF/VI versus BDP/FOR, treatment persistence was greater [hazard ratio, 0.76 (p < 0.0001)].

Overall, patients initiating FF/VI and BDP/FOR had numerically similar exacerbation rates; of the patients continuing 12 months' uninterrupted treatment, the FF/VI group had a lower exacerbation rate versus BDP/FOR. Patients initiating FF/VI were less likely to discontinue treatment than those initiating BDP/FOR.