ICS-formoterol rose sharply after the 2020 guideline shift, but SABA-only care persisted
New Zealand’s 2020 asthma guideline change produced a clear shift in reliever prescribing patterns, but the transition was uneven across the national asthma population. In a retrospective patient-level analysis of 324,972 people aged 12 years and older, ICS-formoterol use expanded after the guideline update, yet a large share of patients remained anchored to SABA-based treatment.
Before the change, the SABA-only category was the most common medication pattern, accounting for 30.5% of the cohort. After the guideline update, the proportion dispensed the preferred anti-inflammatory reliever regimen, ICS-formoterol without SABA, more than doubled from 10.9% to 23.8%. Any ICS-formoterol use climbed from 18.9% to 34.5%.
The direction of travel is clinically important. Airway inflammation, not bronchospasm alone, drives asthma risk, and an anti-inflammatory reliever strategy is designed to treat both the immediate symptoms and the underlying inflammatory flare at the same moment. In practice, that means every reliever dose can carry anti-inflammatory effect rather than providing bronchodilation alone. The prescribing data show that clinicians and patients did move toward that model, but not at the pace implied by the guideline revision.
SABA exposure still dominated much of routine care. Across the full cohort, the proportion dispensed any SABA fell from 74.1% to 60.8%, yet the most stubborn pattern was SABA monotherapy. Of those dispensed SABA only before the guideline change, 57.1% remained on SABA only afterward. A smaller share moved into the recommended reliever pathway: 17.6% shifted to ICS-formoterol without SABA, and 8.1% shifted to ICS-formoterol with SABA.
Maintenance users switched more readily than patients starting from SABA alone
The clearest uptake occurred in people already receiving maintenance therapy. Continuity was high among those dispensed ICS, with 55.9% remaining in that category, and even higher among patients already taking non-formoterol ICS-LABA, where 74.3% continued on the same regimen. Continuity was highest of all among those already dispensed ICS-formoterol, where 89.0% stayed within that treatment class.
This pattern fits prescribing behavior at the bedside. Moving a patient already engaged in inhaled therapy to a preferred regimen is operationally simpler than converting someone whose asthma care has been built around a single bronchodilator inhaler. The latter group often has lower treatment intensity, fewer scheduled review points, and more opportunities for therapy to remain unchanged despite evolving guidelines. The data show that the guideline shift was absorbed most efficiently where a maintenance inhaler framework already existed.
Among patients dispensed ICS-formoterol after the guideline update, 31.1% were also dispensed a SABA. Co-dispensing was even more common in the comparator maintenance groups: 63.8% of non-formoterol ICS-LABA users and 70.2% of ICS users also received a SABA. In other words, even when anti-inflammatory control therapy was in place, rescue bronchodilator prescribing remained deeply embedded in routine practice.
That co-prescribing pattern matters because it signals incomplete replacement of the older SABA-centered model. The AIR approach is meant to reduce reliance on bronchodilator-only treatment, but overlapping prescriptions suggest that many patients may have been given the new regimen without fully retiring the old one. In a disease where poor adherence and fragmented inhaler use are common, the presence of multiple inhalers can dilute the intended simplicity of the updated strategy.
Difference-in-differences confirmed a selective policy effect, not a full population reset
Children younger than 12 years, who were outside the scope of the recommendation, served as a comparison cohort for difference-in-differences estimation. Using that reference group, the estimate for ICS-formoterol without SABA was +8.4 percentage points, and total SABA supply fell by 8.3 percentage points. The estimate for SABA monotherapy, however, was only -0.2 percentage points.
The contrast is revealing. Prescribing changed enough to increase use of the preferred anti-inflammatory reliever and reduce overall SABA supply, but it did not produce a meaningful collapse in SABA-only treatment. The guideline reached into prescribing habits, yet it left the most entrenched pattern largely intact. That gap is not a statistical footnote; it is the central implementation problem exposed by the study.
Because the study used national linked health datasets, it captures real-world practice across the entire asthma population rather than a selected clinic sample. The retrospective design also makes the findings especially relevant to health-system implementation, since the results reflect what happened after publication of the guideline rather than what clinicians intended to do in a controlled setting. The post-guideline period therefore acts as a practical test of whether evidence translated into routine care at scale.
The answer is mixed. The policy shift was substantial, but selective. Patients already receiving maintenance inhalers were much more likely to stay within or move toward anti-inflammatory therapy, while patients on SABA-only regimens were far less likely to switch to the preferred approach. For national asthma control, that leaves the largest implementation burden exactly where the therapeutic need is greatest: among people who continue to rely on bronchodilator-only treatment despite modern guidance.
Further targeted strategies beyond guideline publication will be needed to close that gap. Simply changing the written recommendation was not enough to displace entrenched SABA monotherapy across the country, and the remaining question is how best to convert that recommendation into consistent, patient-level prescribing change in routine care.