It is good to see any cluster randomised
It is good to see any cluster-randomised SCR7 trial of a handwashing intervention, much more a successful one. Globally there is considerable funding invested in handwashing promotion, yet the kind of rigorous study published in this month\'s issue of by Adam Biran and colleagues is rare in the literature. As rare are the implementation details provided, which are often lacking from publications focused on results. Yet such details are important for those seeking to replicate or scale up the intervention. Assessment of study outcomes through structured observation is another laudable aspect of the study method. However, we wonder whether the mechanism of action of the behavioural intervention was different from the mechanism assumed by Biran and colleagues. Although it is plausible that the intervention acted on the intended emotional drivers of handwashing (nurture, disgust, affiliation, status), which in turn resulted in increased uptake of handwashing, these determinants were not measured. Members of Biran and colleagues\' research team have previously systematically measured these determinants in Kenya. Yet it is possible that the changes observed in this study occurred through a different pathway. An alternative explanation would be that the implementing team created ideal conditions for formation of new habits: a break in the routine, cues to perform the new behaviour, promotion of new social norms, creation of a stable context for practising the behaviour, and exhortation to practise the new behaviour repeatedly. The break in the routine and promotion of repeated practice were created by community and school events, and followed by household visits. The stable context for practising the behaviour was evidenced by the fact that every household had soap at baseline, more than 90% of households had access to water within their yards or nearby, and this availability was enhanced by installation of handwashing stations in schools and their monitoring before eating. New social norms were promoted in multiple ways including modelling by village chairmen and group pledgings, and might have persisted beyond the end of the long and short intervention, thus explaining the surprisingly sustained effect over time. Given these favourable conditions for adoption of new habits created by the intervention team, thigmotropism is possible that significant increases in handwashing uptake would have been achieved even in the absence of message content related to nurture and disgust. Biran and colleagues contend that message content related to “rational health beliefs” does not effectively increase handwashing practice, and indeed this is also long-standing concern in HIV prevention, although rational health beliefs do appear to be beneficial for smoking. A meta-analysis of the effectiveness of constructs in the Health Belief Model in predicting behaviour found that “the relationship between susceptibility beliefs and behaviour was almost always near zero”. This finding, combined with some of Biran and colleagues\' previous work on disgust, strengthens the rationale for a focus on emotions (nurture and disgust) in this study. Interestingly, the previous study indicated a significant effect of disgust on handwashing for men but not women. It would be illuminating in a future study to have two different intervention groups: one with the optimum conditions for formation of new habits mentioned above but without the content related to emotions (nurture and disgust), and the other with the full intervention that was tested. So although the argument for disgust as a key motivator for handwashing is intriguing, further research is needed to provide definitive evidence of its effectiveness. We view affiliation, social norms, and status as motives differing from emotions, and are unclear whether they are core components of a package of emotion-related interventions. Finally, the level of handwashing uptake achieved for key occasions post-intervention was comparable to that of other studies—eg, by Huda and colleagues—and might not be high enough to have an effect on public health. Creation of a more enabling environment by means of multiple conveniently placed and replenished handwashing stations in and around the home might be needed to achieve a higher, more effective, increase in handwashing with soap at key occasions.