Feasibility of an incentive scheme to promote active travel to school: a pilot cluster randomised trial

Samuel Ginja, Bronia Arnott, Vera Araujo-Soares, Anil Namdeo, Elaine McColl

In Great Britain, 19% of trips to primary school within 1 mile, and 62% within 1-2 miles, are by car. Active travel to school (ATS) offers a potential source of moderate-to-vigorous physical activity (MVPA). This study tested the feasibility of an intervention to promote ATS in 9-10 year olds and associated trial procedures. A parallel cluster randomised pilot trial was conducted over 9 weeks in two schools from a low-income area in northeast England. Measures included daily parental ATS reports (optionally by SMS) and child ATS reports, as well as accelerometry (ActiGraph GT3X+). At baseline, all children were asked to wear the accelerometer for the same week; in the post-randomisation phase, small subsamples were monitored each week. In the 2 weeks when a child wore the accelerometer, parents also reported the start and finish times of the journey to school. The intervention consisted of a lottery-based incentive scheme; every ATS day reported by the parent, whether by paper or SMS, corresponded to one ticket entered into a weekly £5 voucher draw. Before each draw session, the researcher prepared the tickets and placed them into an opaque bag, from which one was randomly picked by the teacher at the draw session. Four schools replied positively (3.3%, N = 123) and 29 participants were recruited in the two schools selected (33.0%, N = 88). Participant retention was 93.1%. Most materials were returned on time: accelerometers (81.9%), parental reports (82.1%) and child reports (97.9%). Draw sessions lasted on average 15.9 min (IQR 10-20) and overall session attendance was 94.5%. Parent-child report agreement regarding ATS was moderate (k = 0.53, CI 95% 0.45; 0.60). Differences in minutes of accelerometer-assessed MVPA between parent-reported ATS and non-ATS trips were assessed during two timeframes: during the journey to school based on the times reported by the parent (U = 390.5, p < 0.05, 2.46 (n = 99) vs 0.76 (n = 13)) and in the hour before classes (U = 665.5, p < 0.05, 4.99 (n = 104) vs 2.55 (n = 19)). Differences in MVPA minutes between child-reported ATS and non-ATS trips were also significant for each of the timeframes considered (U = 596.5, p < 0.05, 2.40 (n = 128) vs 0.81 (n = 15) and U = 955.0, p < 0.05, 4.99 (n = 146) vs 2.59 (n = 20), respectively). Data suggest the feasibility of an ATS incentive scheme and of most trial procedures. School recruitment stood out as requiring further piloting. ClinicalTrials.gov: NCT02282631. Registered 5th September 2014.