Aims

Varicella zoster virus (VZV) is highly infectious and initially presents as varicella (chickenpox, CP); following recovery VZV poses a lifetime risk of re-emergence as herpes zoster (shingles, HZ). At the time of this study, the Australian National Immunisation Program funded a single dose of VZV vaccine as MMRV (measles, mumps, rubella, varicella) at 18 months of age, and a HZ vaccine for persons aged 70 to 79 years. However, numerous studies have concluded that a two-dose VZV regimen during childhood may be more effective and that the optimal timing for HZ vaccination is uncertain. The aim of this paper was to describe the development and validation of an agent-based model designed to evaluate the cost-effectiveness of various varicella vaccination strategies in Australia.

Methods

During calibration, model parameter values were similar to an earlier published model, with strong agreement between modelled and published CP seropositivity and HZ incidence. Modelling an unvaccinated population, the model displayed a steady state incidence of CP and HZ, with the associated hospitalisations, deaths, and costs in alignment with published values. In the vaccination scenario, the model predicted the marked reduction in CP incidence and characteristic initial increase in HZ reported in published studies.

Results

During calibration, model parameter values were similar to an earlier published model, with strong agreement between modelled and published CP seropositivity and HZ incidence. Modelling an unvaccinated population, the model displayed a steady state incidence of CP and HZ, with the associated hospitalisations, deaths, and costs in alignment with published values. In the vaccination scenario, the model predicted the marked reduction in CP incidence and characteristic intial increase in HZ reported in published studies.

Conclusions

This agent-based model can be used to provide cost-effectiveness estimates for various VZV vaccination strategies in an Australian context.