Comparative adherence and persistence of single-inhaler and multiple-inhaler triple therapies among patients with chronic obstructive pulmonary disease in Japan: a retrospective cohort study

Abstract

Objectives: To evaluate and compare medication adherence and persistence for patients newly initiating single-inhaler triple therapy (SITT) and multiple-inhaler triple therapy (MITT) for chronic obstructive pulmonary disease (COPD) in Japan.

Design: Retrospective, new-user, active comparator, observational cohort study using inverse probability of treatment weighting.

Setting: Health insurance claims data from the Medical Data Vision Co., Ltd, hospital claims database.

Participants: Adults diagnosed with COPD at age ≥40 years newly initiating MITT or SITT (fluticasone furoate/umeclidinium/vilanterol (FF/UMEC/VI) or formoterol fumarate/budesonide/glycopyrronium) from 1 September 2019 to 31 July 2021.

Primary and secondary outcome measures: The primary outcome was medication adherence compared between patients using SITT and MITT, assessed by the proportion of days covered ≥80%. Secondary outcomes included medication persistence (time from index treatment initiation to discontinuation) compared between patients using SITT and MITT and medication adherence compared before and after the switch in a subgroup of patients switching from MITT to SITT.

Results: We included 2575 MITT and 2962 SITT users with similar baseline characteristics following weighting. The proportion of adherent patients was significantly greater for SITT versus MITT users at 6 months (19.7% vs 10.2%, p<0.0001), 12 months (6.0% vs 3.8%, p=0.0009) and 18 months (3.8% vs 1.4%, p<0.0001) post-index. Median persistence was also significantly higher for SITT versus MITT users (2.0 vs 1.0 months, p<0.001). Comparing specific SITT versus MITT, the proportion of adherent patients at each time point and median persistence was greater for FF/UMEC/VI. In patients switching from MITT to SITT (n=688), the proportion of adherent patients increased postswitch at the class level and for FF/UMEC/VI specifically.

Conclusions: Patients with COPD in Japan who were newly initiating SITT had greater medication adherence and persistence compared with those on MITT up to 18 months following initiation.

Keywords: Drug Combinations; Medication Adherence; Medication Persistence; Pulmonary Disease, Chronic Obstructive; Respiratory Therapy.

Figure 1. (a) SITT (overall) versus MITT adherence at 6, 12 and 18 months (PDC ≥80%). (b) SITT (overall) versus MITT ORs for medication adherence at 6, 12 and 18 months (PDC ≥80%). (c) SITT (overall) versus MITT persistence (14-day permissible gap). ORs, 95% CIs and p values were generated from a logistic regression model. Kaplan-Meier persistence rates were evaluated for up to 18 months following the index date and compared between cohorts using Cox proportional hazards models. MITT, multiple-inhaler triple therapy; PDC, proportion of days covered; SITT, single-inhaler triple therapy.

Figure 2. (a) FF/UMEC/VI versus MITT adherence at 6, 12 and 18 months (PDC ≥80%). (b) FF/UMEC/VI versus MITT ORs for medication adherence at 6, 12 and 18 months (PDC ≥80%). (c) FF/UMEC/VI versus MITT persistence (14-day permissible gap). ORs, 95% CIs and p values were generated from a logistic regression model. Kaplan-Meier persistence rates were evaluated for up to 18 months following the index date and compared between cohorts using Cox proportional hazards models. FF/UMEC/VI, fluticasone furoate, umeclidinium and vilanterol; MITT, multiple-inhaler triple therapy; PDC, proportion of days covered; SITT, single-inhaler triple therapy.

Figure 3. (a) FOR/BUD/GLY versus MITT adherence at 6 and 12 months (PDC ≥80%). (b) FOR/BUD/GLY versus MITT ORs for medication adherence at 6 and 12 months (PDC ≥80%). (c) FOR/BUD/GLY versus MITT persistence (14-day permissible gap). ORs, 95% CIs and p values were generated from a logistic regression model. Kaplan-Meier persistence rates were evaluated for up to 18 months following the index date and compared between cohorts using Cox proportional hazards models. FOR/BUD/GLY, formoterol fumarate, budesonide and glycopyrronium bromide; MITT, multiple-inhaler triple therapy; PDC, proportion of days covered; SITT, single-inhaler triple therapy.

Figure 4. Medication adherence (PDC ≥80%) following switch from MITT to SITT (overall). (a) MITT to SITT.* (b) MITT to FF/UMEC/VI.^† (c) MITT to FOR/BUD/GLY.^‡ *For the assessment of adherence pre/post MITT to SITT switch, patients were removed from the analysis if they switched back to the comparator (ie, MITT). ^†For assessment of adherence pre/post MITT to FF/UMEC/VI switch, patients were removed from the analysis if they switched back to the comparator (ie, MITT) OR competitor (ie, FOR/BUD/GLY). ^‡For assessment of adherence pre/post MITT to FOR/BUD/GLY switch, patients were removed from the analysis if they switched back to the comparator (ie, MITT) OR competitor (ie, FF/UMEC/VI) and switch at 18 months is not reported due to low sample size. FF/UMEC/VI, fluticasone furoate, umeclidinium and vilanterol; FOR/BUD/GLY, formoterol fumarate, budesonide and glycopyrronium bromide; MITT, multiple-inhaler triple therapy; PDC, proportion of days covered; SITT, single-inhaler triple therapy.