The educational resource was acceptable to schools and delivered in most intervention schools (85%) within the trial but few control schools (38%) outside the trial setting. Delivery of the intervention was also more likely to occur in schools who participated in the sub-study, where additional student absence data were collected, and university staff had more opportunities for contact with these schools. Fidelity of delivery was also reportedly better in intervention schools than control schools in the trial. These results suggest that delivery of the intervention beyond the trial is unlikely to be sustainable using this model of a centralised, non-research agency to coordinate intervention delivery. Some variations in dose were also apparent and optional intervention activities that might have helped to reinforce hand washing messages, particularly at home, were rarely used. Similar variations in fidelity and dose have been observed in other studies of health promotion interventions in school settings. A review of drug abuse prevention programs taught by teachers found considerable variability in the number of key curriculum objectives covered or the number of modules taught
. Teachers delivering a nutrition curriculum completed 70% of the lessons with a rating of 76% faithful to the curriculum
. Reach in the Trial of Activity for Adolescent Girls was high, with 91% of girls in seventh-grade and 77% of girls in eighth grade taught all of the health education lessons
. The level of fidelity was found to be acceptable in 76% and 64% of these lessons in the first and second year of the study, respectively
. Whether the dose and fidelity obtained in the current study were sufficient to bring about the level of behaviour change required to reduce infection-related absence will only be able to be inferred from the main trial results. The reach obtained in intervention schools, however, compares favourably with a review of prevention and health promotion programs for children and adolescents where few studies documented implementation levels greater than 80%
. Few programs are also able to achieve complete implementation in real-world settings
Qualitative data from this process evaluation provide useful insights for updating the educational resource if it were to be distributed more widely. Teachers generally found the intervention useful and straightforward to deliver. The element generating the most positive reaction was the glitter activity in the ‘Healthy hands, healthy school’ lesson, because of the effectiveness of the kinaesthetic aspect in demonstrating how germs spread, and the importance of hand washing. The glitter lesson may possibly be less effective if some children are passive observers rather than being actively involved. Teachers and pupils also enjoyed the DVD animation showing how to wash hands, although there was some suggestion this may be better suited to younger audiences. The stickers and posters were also well used and received. There was some concern that the ‘What are germs?’ lesson worksheets were too difficult, even for older pupils, as insufficient background information was provided for them to be able to answer the questions. In addition, the pictures of germs that could be coloured in by KS1 pupils were considered too childish because they were presented as caricatures rather than realistic, if somewhat magnified, pictures of germs.
While all four main elements of the educational resource (DVD, lesson plans, posters, and stickers) tended to be delivered to all pupils in the four intervention schools observed, the length of time spent on each lesson varied. It is likely that the depth and breadth of information covered, and therefore the understanding reached, may be quite different for a lesson lasting 10 minutes compared to one lasting over an hour. It is also possible that the rates and duration of lesson delivery noted in this part of the process evaluation may be a result, at least in part, of the classes being observed by a researcher. The trial was designed to be pragmatic, aiming for implementation to be as close to how it would be delivered in practice
. Adaptations teachers make so that it is appropriate and relevant to their class may be acceptable
 and may not result in reductions of effectiveness
, although effectiveness of this educational intervention is yet to be established. Given differing views on resources and different abilities of pupils within year levels across schools, future versions of the resource could include a range of lesson plans and teachers could choose those appropriate for their pupils. This will require communication between program developers, who understand about the essential components of an intervention and its effects, and teachers, who understand pupils, pedagogy and school settings, so that prescription and adaptability can be combined for maximum effectiveness
In addition to updating the resource itself, consideration should be given to the dissemination process as HPA staff coordinating the intervention delivery had limited capacity to follow up and respond to schools. While the HPA are seen to have a role in providing hand washing information, the intervention may be more sustainable if it were embedded as part of the curriculum by education departments, and maximising effectiveness of staff such as healthy school coordinators.
A strength of this study is the use of both quantitative and qualitative methods to measure and explore intervention reach, dose, fidelity, acceptability, and sustainability. In addition, the process evaluation was conducted and analysed by a researcher not involved in the intervention delivery and before the outcome evaluation, so interpretation of key process factors likely to affect outcomes were not influenced by prior knowledge of the outcomes
. A limitation is that direct observation of intervention delivery was only possible within the four intervention schools. We were also unable to elicit reasons why the intervention was not delivered among some schools, due at least in part to limitations on the capacity of HPA staff to ensure that schools were using the resource.
Hand washing is a relatively simple, inexpensive and important public health measure
. Teaching primary school children to wash their hands properly and encouraging regular hand washing may not only reduce infection related absenteeism but also help to habitualise this behaviour at an early age. The ‘Hands Up for Max!’ intervention was a deliberately modest educational intervention, designed to be readily integrated into the existing school curriculum, and provided to schools by a public health body whose main contact with schools traditionally occurs when there is an infectious disease outbreak. Schools are an important setting for health promotion but, as this study demonstrates, simply providing carefully designed health educational resources may not be sufficient to ensure their use. The World Health Organization Health Promoting Schools framework
 also suggests that it is insufficient to merely introduce health into the curriculum but indicates that there also needs to be concomitant changes in the school environment and in the wider community. The ‘Hands up for Max!’ intervention had a homework element with the potential to effect change in the wider community through involving parents, but this element was optional and inconsistently used even within the trial. The intervention did not include any changes to the school environment, but these may be crucial in the case of hand washing
. Changes to the school curriculum and environment, plus the wider community, are only likely to be achieved if health and education authorities work together on a regular and continuing basis. This integrated working can create a shared understanding that education can improve health, and that improving health in schools can also improve educational attainment