Table 2 Feasibility and preliminary effectiveness findings for CLASS II

Twelve schools were recruited. Six were randomly allocated to the smoke-free intervention (intervention arm, n = 245 children), and six delivered usual education only (control arm, n = 236 children). Of 481 children who had cotinine levels indicative of second-hand smoke exposure, 450 were followed up (229 intervention arm; 221 control arm). All schools were retained in the study; 89.9% children (206/229) in the intervention arm and 86.8% (192/221) in the control arm provided a saliva sample for cotinine 2 months post-allocation. Mean cotinine at the cluster level was 0.53 ng/ml (standard deviation 0.36) in the intervention arm compared with 1.84 ng/ml (standard deviation 1.49) in the control, a mean difference of – 1.31 ng/ml (95% confidence interval − 2.86, − 0.24).

In summary, it was feasible to recruit, randomise, and retain primary schools and children. This study, though not powered to detect differences in mean cotinine between the two arms, provided estimates to inform the likely effect size for a future trial.