Highlights
– In 2,598 MASK‑air app users with self‑reported asthma and ICS+LABA use, adherence (≥80% of weekly days) was higher for ICS+other LABA (75.1%) than for ICS+formoterol (59.3%).
– Despite lower reported adherence, ICS+formoterol users recorded lower SABA use (median 57.1% vs 71.4% days) and similar symptom control and work productivity.
– Each additional day of ICS+formoterol use in a week was associated with a statistically significant reduction in weekly SABA use (adjusted risk reduction ~4.1% per day); the effect was larger for other ICS+LABA (about 8.2% per day).
Background
Adherence to controller medications remains a central challenge in asthma care. Suboptimal adherence to inhaled corticosteroids (ICS) contributes to poor control, exacerbations, and increased healthcare utilization. Traditional methods to measure adherence (pharmacy refill records, electronic monitors, clinician estimates) have limitations in scale or feasibility in real‑world settings.
Mobile health (mHealth) tools, including symptom and medication diaries, provide an opportunity to collect longitudinal patient‑reported data at scale. The MASK‑air® app is a multilingual, multi‑country mHealth platform that has been used to track symptoms, medication use, and work productivity in allergic rhinitis and asthma populations.
ICS combined with long‑acting β2‑agonists (LABA) are a mainstay of controller therapy for persistent asthma. Formoterol is a LABA with rapid onset of bronchodilation that, when combined with an ICS, is uniquely suitable for both maintenance and as‑needed use (maintenance-and-reliever therapy, MART). International guidance (e.g., GINA) now endorses low‑dose ICS–formoterol as the preferred reliever in many patients because it reduces SABA reliance and lowers exacerbation risk. However, how patients actually adhere to ICS+formoterol versus other ICS+LABA combinations in everyday life — and how that adherence relates to SABA use and symptom control — is incompletely described.
Study design and methods
The analysis used MASK‑air app data collected between 2015 and 2022 across 27 countries. Eligible users self‑reported a diagnosis of asthma and regular use of an ICS+LABA combination. The investigators analyzed “complete weeks” of app reporting (defined by continuous daily entries for a seven‑day period) as the primary observation unit; a sensitivity analysis included weeks with no more than one missing day.
Users were grouped by their reported controller: ICS+formoterol (ICS+F) versus ICS combined with other LABAs (ICS+other LABA). Key variables extracted from app entries included daily controller use, days with reported SABA use, symptom scores, and a composite symptom‑medication score; work productivity measures available in the app were also examined.
Primary comparisons evaluated the proportion of weeks in which controller adherence reached a conventional threshold (use on ≥80% of days). Regression models assessed associations between the number of controller use days per week and the likelihood of SABA use or suboptimal asthma control during the same week, reporting adjusted percentage change in risk per additional day of controller use. Analyses were reported for both the strict complete‑week sample and the sensitivity sample allowing one missing day.
Key results
Population and data availability: Among 2,598 app users who reported ICS+LABA use, 621 users (23.9%) contributed 4,824 complete weeks of data; 866 users (33.3%) contributed weeks with at most one missing day and were included in sensitivity analyses. Users spanned 27 countries, reflecting a broad international sample of motivated app users.
Adherence: Using the ≥80% days/week threshold, adherence was higher in the ICS+other LABA group (75.1% of weeks) compared with the ICS+formoterol group (59.3% of weeks). This difference persisted in sensitivity analyses that allowed one missing day.
SABA use and symptom control: The ICS+other LABA group reported more frequent SABA use (median 71.4% days with SABA per week) than the ICS+formoterol group (median 57.1% days). Despite differences in SABA use and controller adherence, the groups had similar symptom burden and work productivity scores as captured in the app.
Associations between controller use and SABA use: Regression modelling demonstrated that each additional weekly day of ICS+formoterol use was associated with an adjusted 4.1% lower risk of weekly SABA use (95% CI −6.5% to −1.6%; p=0.001). For ICS+other LABA, the magnitude was larger: an adjusted 8.2% lower risk of weekly SABA use per additional controller day (95% CI −11.6% to −5.0%; p<0.001).
Interpretation of effect sizes: The reported percentage reductions indicate that more frequent controller use in the preceding week was associated with proportionally fewer days the patient reported taking SABA. Both controller types showed statistically significant SABA‑sparing effects with greater daily use, though the absolute adherence rates and baseline SABA use differed between groups.
Expert commentary and interpretation
The MASK‑air analysis yields several clinically relevant observations and raises important interpretive questions:
1) Lower self‑reported adherence among ICS+formoterol users does not necessarily signal worse outcomes. In this study, ICS+formoterol users reported lower weekly controller adherence but also lower SABA use and similar symptom control compared with users of other ICS+LABA combinations. One plausible explanation is that some ICS+formoterol users were following an as‑needed or MART strategy endorsed by international guidance; in that context, measuring adherence as “days with controller taken” using a fixed threshold (≥80% of days) does not capture therapeutic adequacy, since symptom‑driven use may be effective and intentional.
2) The lower SABA use in the ICS+formoterol group aligns with randomized and observational evidence that ICS–formoterol used as reliever reduces SABA reliance and exacerbations. However, the data here are patient‑reported and lack prescription context: we do not know whether users were explicitly prescribed MART, fixed maintenance plus reliever, or a separate rescue SABA. App data alone cannot reliably distinguish maintenance versus reliever dosing or capture inhaler technique and dose counts.
3) Measurement and selection bias are important. MASK‑air users represent a self‑selected, digitally engaged cohort. They may have higher health literacy and different behavior from the wider asthma population. Self‑reported medication entries can over‑ or under‑estimate true inhaler use compared with electronic dose counters.
4) The counterintuitive finding that each additional day of ICS+other LABA use produced a numerically larger SABA‑sparing association than ICS+formoterol should be interpreted with caution. Differences may reflect baseline imbalance in SABA use, disease severity, prescribing practice, or user interpretation of how to log as‑needed formoterol doses. Alternatively, the stronger association in the ICS+other LABA group might reflect that these users were more consistently logging maintenance doses (thus reinforcing the adherence metric), whereas ICS+formoterol users could have been logging fewer days but with more potent symptom relief per use.
Limitations
– Self‑reported app data lack verification with pharmacy records, electronic inhaler monitoring, or clinical outcomes such as exacerbation rates and lung function.
– The study cannot distinguish prescribed regimens (MART vs fixed maintenance), actual dose counts, or correct inhaler technique.
– Selection bias: MASK‑air users are a motivated subset and not representative of all patients with asthma.
– Residual confounding: disease severity, comorbidities, and local prescribing patterns could influence both choice of LABA and observed behaviours.
Clinical implications
The study provides real‑world evidence that app‑reported adherence metrics must be interpreted in light of the pharmacology and intended use of the medication. For ICS+formoterol, lower day‑based adherence does not necessarily translate into worse control or higher SABA dependence, likely reflecting differences in treatment strategy (as‑needed ICS–formoterol use is an effective reliever option). Clinicians and digital health designers should account for regimen type when defining adherence thresholds in mHealth tools.
mHealth platforms like MASK‑air can help identify patterns of SABA overuse or underuse of controllers and could be integrated with clinical workflows to prompt review, education, or stepped care. However, linking app entries to prescription records and objective inhaler data would strengthen inferences.
Future research directions
– Integrate app‑based self‑reporting with electronic dose counters and pharmacy dispensing data to validate self‑reported adherence and characterize dose patterns.
– Distinguish users on MART/as‑needed ICS–formoterol regimens from those on fixed maintenance therapy to better align adherence metrics with therapeutic intent.
– Evaluate whether real‑time app feedback or clinician alerts based on SABA overuse can reduce exacerbations and optimize controller use in randomized implementation trials.
Conclusion
In a large international sample of digitally engaged asthma patients, adherence to ICS+LABA as captured by the MASK‑air app was generally high among those who consistently logged data. Users of ICS+formoterol reported lower day‑based adherence by conventional thresholds but demonstrated lower SABA use and comparable symptom control to users of other ICS+LABA combinations. These findings underscore the need to interpret mHealth adherence metrics within the clinical context of the prescribed regimen, and they support the real‑world relevance of ICS–formoterol strategies that reduce reliance on SABAs.
Funding and clinicaltrials.gov
Funding and registration information are reported in the original publication: Sousa‑Pinto B et al., Pulmonology 2025 (see reference). MASK‑air is an investigator‑led digital health program with multiple sources of academic and institutional support; specific funding details should be consulted directly in the original article.
References
1. Sousa‑Pinto B, Louis R, Anto JM, et al. Adherence to inhaled corticosteroids and long‑acting β2‑agonists in asthma: A MASK‑air study. Pulmonology. 2025 Dec 31;31(1):2416869. doi: 10.1016/j.pulmoe.2023.07.004. Epub 2024 Oct 25. PMID: 37543524.
2. Global Initiative for Asthma (GINA). Global Strategy for Asthma Management and Prevention. 2023 update. Available from: https://ginasthma.org/ (accessed 2025).
3. World Health Organization. Adherence to Long‑term Therapies: Evidence for Action. 2003. Available from: https://www.who.int/chp/knowledge/publications/adherence_report/en/ (accessed 2025).
AI image prompt for article thumbnail
A modern smartphone screen showing an asthma app dashboard with daily medication adherence bars, inhaler icons (ICS+LABA and SABA), and symptom diary entries; a stethoscope and a notepad on a clinician’s desk in soft focus background; neutral clinical lighting, professional and patient‑centred tone.
