Using mathematical models as a toolbox for designing trials to prevent transmission of infectious diseases

Abstract Background The COVID-19 pandemic showcased the importance of understanding and measuring the vaccine effects against viral transmission, but clinical trials to study this effect are difficult to design because these studies need to capture all transmission events, regardless of symptoms and are costly and challenging to implement. Here, we use mathematical models of infectious disease transmission as a tool in the early phase of clinical study design to provide insights about the statistical properties of the design. We present an extended example simulating a household study design to evaluate vaccine effects against transmission. Methods We developed a highly detailed agent-based model of SARS-CoV-2 transmission and vaccination under two scenarios: a high and a low background viral transmission. We simulated the implementation of a household study design where primary participants and household members were tested at fixed intervals of time (daily, twice per week or weekly), and vaccine efficacies against transmission, disease given infection, and susceptibility to infection are estimated. We compared the estimated vaccine efficacies with those inputted to the model. Results Our results showed that the household study design was able to recover with great accuracy the vaccine effects against disease or susceptibility, regardless of the frequency of testing. However, for vaccine efficacy against transmission, the frequency of testing and the background incidence level mattered, and only daily testing accurately estimated the vaccine effect against transmission. Conclusions Using mathematical models to simulate a clinical study can help understand the intrinsic and extrinsic factors that will affect the desired estimands.