April 30 2009

On this site we seek to predict the progression of the influenza epidemic that may occur in the UK. Our model of influenza spread uses the UK census data as well as the National Travel Survey to predict how people move on a day-to-day basis. If some people are infected with the virus, we can use this model to predict how the infection might spread spatially from initial sources.

Two confirmed cases were identified on Monday, 21st April 2009, located near Glasgow in Scotland. On Wednesday 29th April (day 3 of the potential epidemic), we have learned that three new cases have been confirmed: one in Devon, one in Redditch (West Midlands) and one in North London.

We use this information as a starting point for our model and estimate how many cases we could expect in the next few days. We consider four separate possible cases:

1. 1)Worst case scenario (R0 ~ 2) The epidemic is allowed to run its natural course, with no intervention from the authorities and no modification of human behaviour. (R0=2 means that at the beginning, and on average, each case infects two subsequent cases, and these two cases would, on average, produce a further four.) This is clearly not going to happen, and is described for comparison with more realistic settings. In this setting, by day 10 we expect to see 105 cases in total. By day 20 of the epidemic we expect to see over 1800, with the total number of infected people reaching the 18 million by day 60. The epidemic would be expected to last around 140 days. For more details on these simulations please see here.
   

2. 2)Weak control measures (R0 ~ 1.5). For a mild reduction in transmission, we would expect to see 70 cases by day 10, 530 cases by day 20 and 6 million by day 60. The poorly controlled epidemic would be expected to last 230 days and affect 33 million people.
   

3. 3)Moderate control measures (R0 ~ 1.25). Strong but insufficient control. By day 10 we would expect to see 54 cases, with the number rising to 234 cases by day 20 and the disease reaching 39 thousand people by day 60. In this case we would expect the disease to persist for over two years (700 days) at a low level in total reaching 21 million people.
   

4. 4)Control measures are just sufficient (R0 ~ 1). Strong and sufficient control of the epidemic. In this case the epidemic is most likely to fade out given just the current initial cases. We would expect it to persist for less than 30 days at a very low level, affecting around 100 people in total. However, should a global pandemic occur, further introductions of the infection are very likely. This would prolong the epidemic in the UK, and the number of infected individuals would rise.

The epidemic of swine flu that has begun spreading from Mexico looks to become a worldwide phenomenon. If we examine the report of numbers of confirmed cases world wide (outside of Mexico, where reporting seems to be more reliable), it appears that we are in early stages of exponential growth world wide. The WHO has raised the alert over swine flu to level 5, indicating human-to-human transmission in at least two countries. This suggests that a world-wide pandemic is possible in the coming months.

We will up-date this graph as the epidemic progresses and more information becomes available.

How can you help? Please take our online contact survey to help us better predict and control the future spread of infectious disease. www.contactsurvey.org.

Page contact l.danon@warwick.ac.uk