Modelling adaptive behaviour to predict the effectiveness of contact-tracing apps integrated in risk mitigation strategies

Abstract

The Covid-19 pandemic has outlined the creativeness of each country in the definition of the most appropriate public health management strategies to curb the spread of the SARS-CoV-2 virus, ranging from total lockdowns to trusting in people’s common sense to act properly. Several non-pharmaceutical health interventions have nationally been issued either on a voluntary bases or required by law. Maintaining 1.5 m distance and wearing protective face masks provide two respective examples. Proximity contact-tracing apps have furthermore been launched at a national scale to alleviate the pressure on the overburdened testing capabilities by means of automatic contact-tracing. Also the Dutch Ministry of Health launched its CoronaMelder on October 10, 2020. However, its efficacy is being questioned with the consideration of its instigated user behaviour. More specifically, it is being hypothesized that the CoronaMelder app instigates users to comply less with the issued non-pharmaceutical health interventions, thereby countering their expected beneficial effects in curbing the spread of the virus. An explorative sequential design provided the answer, where literature review findings were combined to conceptualize the effect of the app on attitude, perception and eventually behaviour. Structural equation modelling performed on data collected through an online survey (N=776) outlined the lack of causality from the CoronaMelder to behaviour and as such, compensatory behaviour was excluded. Moreover, it was found that older and higher educated people were more prone to install the app, similarly to people at risk and those who had previously contracted the virus. Eventually these results were placed in the context of behaviourally-informed politics to assess the management of nationally launched digital contact-tracing apps.