Presented by Dr Miguel Juarez, from the School of Mathematics and Statistics

Recent research estimates the median cost of releasing an intervention to the market just below $1b and taking just over 10 years on average.  If we include the risk of animal and human testing in the equation, there is a strong case for trying to reduce both cost and duration of clinical trials. 

For some years now, there has been a strong drive for using in silico technologies to help with clinical trials. Mechanistic simulations and agent based models, amongst other approaches, have been developed to replicate the behaviour of a specific system in order to create virtual patients, aiming to simulate the effect of interventions.  One of the main obstacles to the implementation of these technologies in frailty, is the lack of a consensus definition of the condition. We advance that probabilistic models are a more suitable approach here, facilitating identification, measurement and propagation of the different sources of uncertainty in identifying and classifying frailty, enabling forecasting of interventions for use in augmented clinical trials.