Melbourne School of Population and Global Health - Theses
Permanent URI for this collection
Search Results
1 - 1 of 1
-
ItemUnderstanding the determinants of pertussis spread to improve herd protection of vulnerable infants( 2014)In the first half of the 20th century, 2-5 yearly whooping cough epidemics caused a substantial burden of death and disease in Australia. The introduction of mass vaccination in the 1950s dramatically reduced disease incidence, but from the 1980s onwards resurgence of cases was observed, placing vulnerable infants at risk. Similarly to other countries with well-established vaccine programs, from 2008-2012 pertussis incidence reached the highest levels in Australia since the implementation of vaccination. I developed an age structured, deterministic, compartmental transmission model to investigate the underlying infection patterns consistent with Australia's observed pertussis epidemiology, using a variety of local data resources. Varying levels of susceptibility to infection were included to account for the possible dependence of susceptibility and infection characteristics on the time since prior infection or vaccination. Biologically plausible parameter ranges were explored using Latin Hypercube Sampling and the model simulated through the pre-vaccine and vaccine eras. Simulation results were filtered to match aspects of pertussis epidemiology about which we can be reasonably certain from long term surveillance data, including persistence of 2-5 year epidemic cycles, relative maintenance of infant protection and declining natural immunity across all age cohorts. Only simulations with natural immunity lasting decades simultaneously reproduced the adult immunity profiles observed in serosurveillance and the substantial impact of vaccination on incidence. Although the initial impact of vaccination was substantial, fluctuations in vaccination coverage through the pertussis vaccine scare of the 1970s were sufficient to allow breakthrough increases in circulation in the model. The long duration of natural immunity required to match observed epidemiology meant that cohorts infected during periods of low coverage were primed to experience a resurgence decades later. In order to replicate the observed sustained, substantial impact of vaccination on infants, the model had to be configured with the primary course delivered as three separate doses, with additional protection against infection acquired with each dose. This finding was consistent with the reduction in infant disease observed across the 2, 4 and 6 month schedule age points. Additionally, an emergent property of the model was that pertussis vaccines induce herd protection of infants. Using my model to project alternative historical vaccination schedules for simulations matching the key characteristics of pertussis in Australia, I found that of the changes made after 1990, the addition of the pre-school booster in 1995 had the largest impact on infection incidence in infants. While the replacement of the 18mth booster with an adolescent dose from 2003 had a limited impact on infants, the direct effect on the 18mth-4yr age group was substantial. More generally, the addition of extra doses to the vaccine schedule increased herd protection in the model, while altering the ages at which doses are given had substantial direct effects but a limited impact on infants. The challenge for policy makers is to design a vaccine program that allows circulation sufficient to maintain immunity in older age groups through boosting while simultaneously protecting infants from such circulation.