Submitted by Helen C. Scott on Thu, 22/02/2018 - 09:01
Epidemics – in plant, animal or human populations – can have devastating consequences. This puts rapidity of response – starting to manage disease as soon as possible – very high on the media and policy agendas. Nevertheless, early action can also be criticised. For example, very costly measures were put in place to counter the perceived threat of a pandemic of H1N1 influenza in 2009. The outbreak eventually turned out to be much less serious than expected. This lead to speculation that the response was disproportionate to the real level of threat. Perhaps waiting – allowing the transmission dynamics to be better characterised – might have been a better strategy?
A recent study by Robin Thompson (ex-PhD student in Plant Sciences), Chris Gilligan and Nik Cunniffe investigates precisely this issue. Their paper in PLOS Computational Biology introduces a new method to determine the optimal time to begin disease management. Their Control Smart Algorithm balances the cost of waiting – which allows the epidemic to become larger – against the benefit, whereby waiting allows transmission to be characterised more precisely, and so management to be targeted more effectively.
By running many millions of simulated outbreaks using a simple model of foot-and-mouth disease spreading through a population of farms, the Control Smart Algorithm has been shown to outperform simpler methods of targeting when and how to manage disease. The algorithm takes as input the type of data that would be available during a real outbreak. The next step is therefore to test the performance of the method on real-time data from an actual emerging outbreak.
Thompson RN, Gilligan CA, Cunniffe NJ. (2018) Control fast or control smart: When should invading pathogens be controlled? PLOS Computational Biology.