O

mg 15 4

Figure 10.3 (a) Observed biweekly measles notifications for London (corrected for under-notification) in the pre-vaccination era (circles). The line shows a corresponding deterministic simulation of a forced SIR model set in a time-series framework (the TSIR model; see Bjornstad etal. (2002) and Grenfell etal. (2002) for details and fitting procedures). The lower reflected graph in (a) shows 10 replicate stochastic simulations of the same model. This underlines that demographic stochasticity has remarkably little effect on the endemic limit-cycle behaviour of measles in large communities. (b) Corresponding annual birth rates (per thousand population) for London; from the 1950s onwards we see sustained biennial cycles; by contrast, the baby boom during the 1940s increased the recruitment of susceptible individuals, driving annual measles epidemics.

0 0

Post a comment