OBJECTIVES: Dynamic transmission models (DTMs) play a crucial role in designing and evaluating the public health impact of vaccination programs. In the case of varicella vaccines, DTMs have been used to address concerns regarding a potential increase in varicella infection age and herpes zoster incidence following the implementation of universal varicella vaccination (UVV) in young children. This study aimed to validate the dynamic modelling predictions using actual data from a country with a long-term UVV program.

METHODS: A newly developed compartmental, age-stratified DTM was utilized to estimate the public health impact of UVV program in Germany, considering the settings and timeframe of the existing UVV program introduced in 2004. Country-specific varicella seroprevalence during the pre-vaccination period, social contact patterns, and disease characteristics were obtained from the literature. The national statistics provided information on the year of vaccine introduction, the strategy employed, and the observed coverage rates. Vaccine characteristics were based on clinical trials. The model-predicted change in varicella incidence was compared to the observed data following vaccination introduction in the country.

RESULTS: The incidence rate predicted by the DTM following UVV introduction in Germany reproduced well the trend observed in reported data, i.e. decrease by over 50% after 4 and by over 80% after 7 years post-vaccination. Validation against the data observed after 10 years from the vaccine introduction showed that in long-term the model tends to be conservative, predicting varicella incidence higher than observed.

CONCLUSIONS: The validation results provide strong support for the credibility of the DTM as a method for assessing the impact of varicella vaccination programs implemented in selected countries. In contrast, a static model using the same settings significantly underestimated the impact of these vaccination programs. However, it is important to acknowledge the influence of population heterogeneity and vaccination settings on the reliability of the model.