A dynamic model of bovine tuberculosis spread and control in Great Britain

Rapid assessment of influenza vaccine effectiveness: analysis of an online survey


Using an online survey of healthcare-seeking behaviour to estimate the magnitude and severity of the 2009 H1N1v influenza epidemic in England 

Epidemiologic inference from the distribution of tuberculosis cases in households in Lima, Peru 

The impact of realistic age structure in simple models of tuberculosis transmission 

Public Health, Epidemics and Maths

Rich kids, poor kids

I first became interested in public health during a year as a medical student at the University of Glasgow. I had this book 'Essential Public Health Medicine' by Donaldson & Donaldson (little did I know at the time that this was Liam Donaldson, to be Chief Medical Officer). In the book they give some statistics about the chances of dying in a fire before the age ten. In the UK, growing in up a low income group, you are 10 times more likely to die in a fire than if you are in a high income group. How can this be? Presumably, poor children aren't intrinsically more flammable than rich kids, are they? No. But maybe their parents are more likely to smoke, or they're less likely to have a fire alarm, or they have more flammable furniture at home, or any number of other reasons that are folded up this ridiculous statistic. 

But rote learning lists of drugs, bones and bits of the lymphatic system was definitely not for me and I switched degree to the brilliant world of Maths for four years. I had a great time thinking about things and almost completely forgot why I had wanted to do Medicine in the first place. At the end of my degree I knew that Maths was for me, but I didn't think that I was quite cut out for pure maths. I took a year off and came across (my future PhD supervisor) Matt Keeling's work on mathematically modelling epidemics.  

Epidemics can be surprisingly predictable. It's amazing, but it's true. Infectious diseases can be modelled with a simple set of differential equations. They might not be able to capture every little blip and bump of an epidemic and they can't predict exactly who will get infected, but they do do a good job of predicting what proportion of people will be affected and how rapidly the epidemic will happen. 

BUT, we can't predict some epidemics. Or, we could improve epidemic models by including the right details so that they can be more accurate in the future. Ahhhh.... but what are the right details? That's exactly what I and others like me are trying to work out.  

Research interests:
bovine TBTuberculosis, Demography, Influenza, Networks