The epidemiology of varicella under a two-dose vaccination strategy: results of a mathematical transmission model with a novel strategy of sensitivity analysis

In 1995 the United States implemented a single-dose strategy of varicella vaccination in infants. Varicella incidence, morbidity, and mortality declined dramatically, though outbreaks continued, even in highly vaccinated populations, and the incidence of varicella began rising in 2003. These events prompted the recommendation of a two-dose vaccination strategy in 2005. In part one of this dissertation, a deterministic, age-structured transmission model of the two-dose strategy is used and predicts a large epidemic of varicella in the near future, even with high second-dose coverage rates. In the long-term, incidence rates under a two-dose regime will be 10% or less of the pre-vaccination rates, compared with up to 50% with a continued one-dose strategy. Varicella cases will consist mostly of mild, breakthrough disease in previously vaccinated individuals. A full sensitivity analysis should be performed on all models of disease transmission. The sensitivity analysis is used to determine the sensitivity of the model output to the values of input parameters, and to the structure of the model itself. In part two of this dissertation, we present a simple, systematic method to perform deterministic sensitivity analysis on a mathematical model of infection transmission, and apply it to the varicella transmission model of part one. The methods are general, and should be applicable to any qualitative transmission model