Once-daily vilanterol-fluticasone versus twice-daily formoterol-fluticasone in chronic obstructive pulmonary diseases: Real-world evidence on dosing frequency, adherence, and clinical outcomes

Study design and setting

This prospective observational study was conducted at a tertiary care center between December 2024 and October 2025, with Institutional Ethics Committee approval and adherence to the Declaration of Helsinki. Patients aged ≥30 years with confirmed COPD Patients aged ≥30 years with confirmed COPD, defined by a post-bronchodilator FEV1/ forced vital capacity (FVC) ratio of <0.70 according to the GOLD 2025 guidelines,^[1] who required LABA/ICS therapy were enrolled after providing written informed consent.

A convenience sample of 120 patients (60/group) was used. Post hoc power analysis demonstrated 82% power to detect a 70 mL FEV1 difference and 78% power to detect a 15%-point adherence difference at α = 0.05. Given the post hoc nature of this calculation, the study is best regarded as a hypothesis-generating pilot.^[13]

Eligibility and treatment allocation

Inclusion criteria comprised GOLD Grade 2–3 COPD (FEV1 30–80% predicted), COPD assessment test (CAT) score ≥10, stable disease, absence of exacerbations in the 4 weeks preceding enrollment for outpatient participants, ability to perform acceptable spirometry, and willingness to purchase prescribed medications and attend follow-up visits. Key exclusion criteria included asthma or asthma-COPD overlap, active tuberculosis, severe cardiovascular disease (NYHA [New York Heart Association] III–IV or recent myocardial infarction [MI]), significant hepatic (Child–Pugh Class B or C) or renal (estimated glomerular filtration rate [eGFR] <30 mL/ min/1.73 m²) impairment, pregnancy or lactation, current use of long-term oxygen therapy exceeding 15 h daily, history of lung resection or lung volume reduction surgery, and recent use of triple therapy or LABA/ICS within 4 weeks.

Treatment decisions were made by physicians based on clinical judgment, patient preference, and economic considerations, reflecting real-world practice. Patients received either VI 25 mcg plus FF 100 mcg once daily via metered-dose inhaler (MDI) or FOR 6 mcg plus FP 250 mcg twice daily through MDI. All medications were purchased by patients at their own expense from local pharmacies, while the institution provided spirometry services free of cost at baseline, 12 weeks, and 24 weeks as part of routine care documentation. Physicians and patients who selected VI/FF were systematically motivated by adherence-related factors, which may independently predict better outcomes.

Assessments and follow-up

Baseline assessment included comprehensive medical history, physical examination, spirometry (performed 15 min after 400 mcg salbutamol via MDI with spacer, per the Indian Chest Society guidelines^[14] using Indian reference equations^[15,16]), CAT, St. George’s respiratory questionnaire (SGRQ), and modified Medical Research Council dyspnea scale. All patients received standardized MDI training and spacer devices. Patients maintained daily symptom diaries recording respiratory symptoms, rescue medication use, adverse events, and missed doses. Follow-up visits at 12 and 24 weeks (±1 week) included pre-dose trough spirometry (22–26 h post VI/FF or 10–14 h post FOR/FP dose), CAT, SGRQ, exacerbation documentation, rescue medication review, inhaler technique verification, and structured adherence assessment through dose counter review, prescription refill records, and patient interview, specifically probing reasons for missed doses. Good adherence was defined as ≥80% of prescribed doses taken.^[17]

Statistical analysis

Data were analyzed using MS Excel and Statistical Package for Social Sciences version 25.0 (significance threshold p <0.05 for primary outcomes). Continuous variables were presented as mean ± standard deviation (SD) or median (interquartile range); categorical variables as frequencies with percentages. Normality was assessed using the Shapiro–Wilk test and Q-Q plots. Primary efficacy analyses used the per-protocol population (patients completing 24-week follow-up with ≥80% adherence); safety analyses included all patients receiving at least one dose. Intention-to-treat sensitivity analyses included all enrolled patients.

Baseline comparisons used an independent t-test, Chi-square, or Fisher’s exact test as appropriate. Within-group FEV1 changes were analyzed by paired t-test; between-group comparisons by analysis of covariance (ANCOVA), adjusting for baseline FEV1, age, sex, smoking status, and COPD severity. Repeated measures analysis of variance assessed treatment-by-time interaction. Exacerbation rates were analyzed by negative binomial regression (adjusted for baseline characteristics) and time-to-first-exacerbation by Cox proportional hazards regression. CAT score changes were analyzed similarly to FEV1.

Given the observational design, propensity score matching (1:1, caliper 0.2 SD) and multivariable adjustment were performed for all primary analyses. The propensity model c-statistic was 0.58, indicating modest discrimination; results should be considered exploratory and do not support causal inference. Multiple imputation (20 imputations) addressed missing data. Bonferroni correction (α = 0.017) was applied for three co-primary outcomes. Pre-specified mediation analysis using the product-of-coefficients method (1000 bootstrap iterations) estimated the proportion of the exacerbation benefit mediated through improved adherence.

Patient enrollment and baseline characteristics

Of 156 screened patients, 120 were enrolled (60/group), and 36 were excluded due to failure to meet inclusion criteria, declined participation, or presence of exclusion criteria. Physician preference drove VI/FF selection in 45 cases (primarily for adherence reasons) and FOR/FP in 38 cases (primarily for cost); patient preference determined allocation in the remainder. Baseline characteristics were well-balanced as illustrated in Table 1; mean age 61.4 ± 8.7 versus 62.1 ± 9.2 years, 78.3% versus 75.0% male, pack-years 32.5 ± 15.8 versus 34.2 ± 16.4, FEV1% predicted 54.2 ± 12.4 versus 53.6 ± 13.1, and CAT scores 21.8 ± 5.6 versus 22.3 ± 5.9 for VI/ FF and FOR/FP, respectively. After propensity matching, 52 pairs were identified with good covariate balance.

Study completion and treatment adherence

Study completion was 95.0% (VI/FF) and 90.0% (FOR/FP). Treatment adherence, the primary focus of this study, was significantly superior with VI/FF [Table 2]. At 24 weeks, 91.3% of VI/FF patients achieved ≥80% adherence versus 76.4% of FOR/FP patients (p <0.001); mean dose counter adherence was 93.7 ± 8.2% versus 81.5 ± 14.6%. Structured interviews identified dosing frequency as the dominant barrier in the FOR/FP group: 33.3% cited forgetting the second dose, 22.2% reported inconvenience or routine disruption, and 11.1% were away from home at dosing time. No VI/FF patient cited inconvenience as a barrier.

Primary outcomes

Both groups improved significantly in trough FEV1 from baseline, with VI/FF showing superior gains [Figure 1]. At 24 weeks, VI/FF showed a mean improvement of 198 mL (14.0%) versus 109 mL (7.8%) with FOR/FP. Between-group comparison revealed an unadjusted difference of 89 mL (95% confidence interval [CI] 52–126, p <0.001) with Cohen’s d of 0.76, an ANCOVA-adjusted difference of 92 mL (95% CI 56–128, p <0.001), and a propensity score-adjusted difference of 87 mL, all highly significant. This difference, while statistically robust, falls slightly below the established MCID (minimal clinically important difference) of 100 mL.^[18] Repeated measures analysis confirmed divergent treatment trajectories with significant treatment-by-time interaction (F = 9.73, p <0.001). Subgroup analyses for exacerbation outcomes showed consistent treatment effects across demographic and clinical subgroups, including age, sex, smoking status, and COPD severity [Table 3]. No significant treatment-by-subgroup interactions were detected (all p-interaction >0.05).

Close

Figure 1: Change in trough forced expiratory volume in 1 s from baseline to week 12 and week 24 data presented as mean FEV1 values. Error bars represent standard error. p-values indicate between-group comparisons at each time point. FOR/FP: Formoterolfluticasone propionate (twice daily), VI/FF: Vilanterol-fluticasone furoate (once daily).

Export to PPT

Exacerbation rates were 0.74 per patient-year for VI/ FF versus 1.37 for FOR/FP [Figure 2]. The proportion experiencing ≥1 exacerbation was 28.1% versus 48.1% (p = 0.028; NNT [number needed to treat] = 5). Cox proportional hazards analysis demonstrated significantly longer time to first exacerbation with VI/FF (hazard ratio 0.52, 95% CI 0.30–0.92, p = 0.021). Adjusted rate ratio on negative binomial regression was 0.61 (95% CI 0.40–0.93, p = 0.029); propensity-matched rate ratio was 0.58 (95% CI 0.38–0.88, p = 0.011). Mediation analysis estimated adherence accounted for ~31% of the exacerbation benefit (indirect effect relative risk (RR) 0.85, 95% bootstrapped CI: 0.76–0.94), suggesting dosing simplification itself is a meaningful clinical intervention, while direct pharmacological effects and unmeasured confounders explain the remainder.

Close

Figure 2: Exacerbation rates by treatment group over 24 weeks, exacerbation rates expressed as events per patient-year. Error bars represent standard error. VI/FF: Vilanterol-fluticasone furoate, FOR/FP: Formoterol-fluticasone propionate.

Export to PPT

CAT scores improved significantly in both groups [Figure 3, Table 4]. Mean reduction at 24 weeks was 6.7 points (VI/ FF) versus 3.9 points (FOR/FP); adjusted between-group difference 2.9 points (95% CI −4.6–−1.2, p = 0.001), exceeding the MCID of 2 points.^[19] Responder rates were 84.2% versus 68.5% (p = 0.046; NNT = 7).

Close

Figure 3: Change in COPD assessment test score from baseline at week 12 and week 24, respectively. Lower scores indicate better symptom control. Error bars represent standard error. CAT: COPD assessment test, COPD: Chronic obstructive pulmonary disease, FOR/FP: Formoterolfluticasone propionate, VI/FF: Vilanterol-fluticasone furoate.

Export to PPT

Secondary outcomes and safety

SGRQ improvements numerically favoured VI/FF (7.8 vs. 4.8 points) but did not reach statistical significance; the 3.0-point difference approaches but does not reach the MCID of 4 points [Table 4].^[20] Rescue medication use was significantly lower with VI/FF (9.2 vs. 12.9 puffs/week; adjusted difference −3.4 puffs/week, 95% CI −5.8–−1.0, p = 0.006).

Safety profiles were comparable. Adverse event rates were 46.7% (VI/FF) versus 53.3% (FOR/FP); serious adverse events were 3.3% versus 6.7% [Table 5]. No deaths or major cardiovascular events occurred in either group.

Strengths and limitations of the study

This study’s key strengths include its prospective real-world design, structured multi-method adherence assessment with qualitative characterization of adherence barriers, formal mediation analysis, and propensity score adjustment, providing evidence specifically relevant to Indian clinical practice that RCTs cannot replicate. However, the non-randomized allocation introduces confounding by indication that statistical methods only partially address; the c-statistic of 0.58 confirms substantial residual confounding, and causal inference is not possible. The modest sample size limits power for secondary endpoints, 24-week follow-up may not capture long-term patterns, and self-reported adherence may overestimate actual intake. The single tertiary center and self-funded medication requirement limit generalizability. These findings should be regarded as hypothesis-generating, warranting confirmation through randomized pragmatic trials with adherence as a co-primary endpoint.