Speaker
Description
Mathematical modelling allows us to answer “if this then what” questions. For infectious disease epidemiology this has been widely used to guide interventions: if we intervene like this, how much disease will we prevent? Also important, though much less widely practised, is the use of mathematical modelling to guide the design of studies and trials: if we measure like this, how much will we learn? Vaccine trials measure the effectiveness of our most important infection prevention tool. Understanding them quantitatively requires modelling the underlying epidemic process and modelling the trial operating on it, both of which may be complex. I will introduce our group’s PRESTO project—“PREpare using Simulated Trial Optimisation”—creating a modelling platform for quantifying the variety of factors on which vaccine trial success depends. These include the pathogen epidemiology, the background trial population and the process for enrolling participants from this, choices of endpoint and testing, logistical and ethical constraints, other interventions in place, and statistical analysis methods. I will focus on a case study implemented in the PRESTO platform, studying Lassa fever. Lassa virus is a WHO and CEPI priority pathogen, with no licensed vaccine but four candidates currently in development. The virus is primarily acquired through exposure to rodents and is endemic in several countries in West Africa. Past infection levels have been observed to predict incidence, unsurprisingly, which naively suggests that locations known historically for highest exposure levels will be favourable locations for trials. However, historical exposure implies pre-existing immunity, which likely limits the additional benefit of vaccination. We quantified the trade-off between these factors, which are more broadly relevant to vaccine trials for endemic diseases, and explored how different enrolment, testing and statistical strategies combine to optimise trial power.
| Expedited Notification | No thanks, I do not require Expedited Notification |
|---|
