Process evaluations of a web-based intervention to increase handwashing during a pandemic: Moving Germ Defence from a randomised controlled trial to public dissemination.

Miller, Sascha; Ainsworth, Ben; Weal, Mark; Smith, Peter; Little, Paul; Yardley, Lucy; Morrison, Leanne

BACKGROUND Washing hands helps prevent transmission of seasonal and pandemic respiratory viruses. The PRIMIT study developed a fully automated, digital intervention to promote handwashing. In a randomised controlled trial during the Swine Flu outbreak, participants who had access to the intervention reported washing their hands more and experienced less respiratory tract infections than those without access. Using these findings, the intervention was subsequently adapted, renamed 'Germ Defence', and a study designed to assess a preliminary dissemination of the intervention to the general public to help prevent the spread of seasonal colds and flu. OBJECTIVE This paper compares process evaluations of the PRIMIT trial and Germ Defence dissemination to examine: 1) how online research enrollment procedures impacted on who used the intervention; 2) intervention usage in the two contexts; 3) whether increased intentions to wash hands are replicated once disseminated. The purpose of this paper is to provide insight so that the Germ Defence intervention could be optimized for wide-scale dissemination in the event of a global pandemic. METHODS The PRIMIT trial ran between 2010 and 2012 recruiting participants offline from General Practices, with restricted access to the intervention (N=9155). Germ Defence was disseminated as an open access website for use by the general public from 2016 to 2019 (N=624). The process evaluation plan was developed using Medical Research Council guidance and the framework for Analysing and Measuring Usage and Engagement Data. Both interventions contained a goal-setting section where users self-reported current and intended handwashing behaviour across seven situations. RESULTS During online enrolment, 54.30% (n=17,511) of PRIMIT study participants dropped out of the study compared to 36.46% (n=358) of Germ Defence users. Having reached the intervention, 93.79% (n=8586) of PRIMIT users completed the core section, whereas 65.06% (n=406) of Germ Defence users reached the same point. Users across both studies selected to increase their handwashing in five out of seven situations, including before eating snacks (PRIMIT MD=1.040 [CI 1.016, 1.063], Germ Defence MD=.949 [CI 766, 1.132]) and after blowing their nose, sneezing or coughing (PRIMIT MD=.995 [CI .972, 1.019], Germ Defence MD=.842 [CI .675, 1.008]). CONCLUSIONS By comparing a preliminary dissemination of Germ Defence to the PRIMIT trial we have been able to examine the potential effects of research procedures on uptake and attrition, such as the sizeable dropout during the PRIMIT trial enrolment procedure that may have led to a more motivated sample. The Germ Defence study highlighted points of attrition within the intervention. Despite sample bias in the trial context, the intervention replicated increases in intentions to handwash when used 'in the wild'. This preliminary dissemination study informed the adaptation of the intervention for the COVID-19 health emergency, and it has now been disseminated globally. CLINICALTRIAL ISRCTn75058295.