Respiratory infections like influenza and SARS-CoV-2 pose significant global health risks due to their high transmissibility and severity. SARS-CoV-2 has caused over 7 million deaths worldwide, and the Lancet Commission estimates a >20% chance of a similar respiratory virus pandemic within a decade. Schools, often poorly ventilated, are high-risk settings for transmission. While COVID-19 school closures may have reduced transmission, they likely caused learning loss, mental health issues and increased burdens on parents and caregivers. Air purifiers with High Efficiency Particulate Air (HEPA) filters may offer a non-disruptive mitigation strategy, but the evidence to support their effectiveness in reducing viral transmission is weak. This protocol describes a cluster-randomised, parallel, two-arm, group sequential superiority trial with an interim analysis—to allow early stopping for efficacy or futility—to estimate the effect of portable air purifiers with HEPA filters in primary school classrooms on student absenteeism.

This group sequential trial will randomise schools (clusters) 1:1 to intervention or control arms in two stages: winter 2025/2026 (N=32 schools; ~736 students) and winter 2026/2027 (N=30 schools; ~690 students). The study setting is Norway. Eligible schools must have classrooms suitable for portable air purifiers, >10 students in grades 5–7 (typically aged 10–13 years) and principal consent. Intervention and control classrooms will each receive two portable air purifiers with HEPA filters operating at a performance equivalent to 3.0 and 0.3 air changes per hour, respectively, with control purifiers acting as shams. Outcomes will be measured during and at the end of a 12-week period. The primary outcome is student absenteeism, measured as full child days of absence aggregated at the class level. An interim analysis is planned at the end of the first stage, with error-spending O’Brien-Fleming stopping boundaries that are binding for efficacy and non-binding for futility. The primary estimand is the marginal incidence rate ratio of student absences, estimated using generalised estimating equations with a negative binomial model to account for overdispersion. Prespecified stopping boundaries will determine stopping, with efficacy boundaries being binding. Treatment effects will be estimated using cluster-bootstrapped CIs adjusted to provide strong control on overall type I and II error probabilities, and a bias correction will be applied if the trial is stopped early for efficacy. All analyses will follow the intention-to-treat principle. The primary question is whether installing and operating air purifiers with HEPA filters (intervention) reduces student absenteeism due to sickness (primary outcome) compared with sham air purifiers (control). Secondary questions examine whether the intervention reduces teacher absenteeism due to respiratory infections, rate and 12-week risk of self-reported respiratory infections among teachers, and teachers’ perceptions of air quality, compared with sham air purifiers. If the trial estimates a statistically significant effect for the primary outcome, a cost-consequence analysis will evaluate the direct and indirect costs associated with operating air purifiers against the potential benefits of reduced student and teacher absenteeism. A process evaluation will explore mechanisms of effect.

The trial has been approved by the Regional Committees for Medical and Health Research Ethics and the National Research Ethics Committee (ref. 737650). Absence data, routinely collected and fully anonymised at the class level, will serve as the primary outcome and does not require informed consent. Informed consent will be obtained from teachers participating in weekly surveys or interviews. Results will be disseminated to stakeholders, participants and the public through peer-reviewed journals, scientific meetings and social media.

NCT07119333.

To explore participants’ experiences of potential adverse effects of the Informed Health Choices secondary school intervention across three trial sites and to revise a framework of potential adverse effects of interventions to improve critical thinking about health choices.

This was a qualitative study. We extracted and analysed relevant data from separate process evaluations in each country. Data came from surveying teachers, observing lessons and group and individual interviews with students, teachers and other stakeholders. We modified and applied framework analysis, including five stages: (1) development of an initial framework of potential adverse effects, (2) familiarisation with the data, (3) indexing, (4) abstraction and synthesis and (5) revising the framework. We applied reflexive strategies individually and as a team.

Lower secondary school in five randomly sampled subcounties of Kisumu County in Kenya, districts representing all five provinces in Rwanda, and six districts in the central region of Uganda, between 2022 and 2024.

Students and teachers in the intervention arms of the trials, parents of students in the intervention arms and administrators at intervention schools, as well as curriculum developers and policy-makers.

The intervention involved providing teachers with a 2–3-day training workshop, and digital classroom resources, including lesson plans for 10 lessons to be delivered over the course of one semester.

We generated findings about potential increases in adverse misunderstandings, anxiety related to transfer of learning, adversely experienced cognitive dissonance, work or schoolwork-related stress, inequity, conflicts and waste. The revised framework includes the same categories of potential adverse effects as our initial framework: decision-making harms, psychological harms, equity harms, group and social harms, waste and other harms. We revised other elements of the framework, including definitions of the categories and its structure.

This study provides insight into the potential adverse effects of interventions to improve critical thinking about health choices. The findings complement those of the trials and country-level process evaluations.