ISPOR Europe 2018
Barcelona, Spain
November, 2018
PRM137
Infection all, Neurological Disorders-all
Research on Methods (RM)
Modeling Methods (MS)
MODELING THE DISTRIBUTION OF TBEV-INFECTED TICKS IN GERMANY TO ESTIMATE TBE-INFECTIONS IN HUMANS
Brugger K1, Schiffner-Rohe J2, Walter M1, Vogelgesang J1, Schöler J3, Rubel F1
1Veterinaermedizinische Universitaet Wien, Wien, Austria, 2Pfizer Deutschland GmbH, Berlin, Germany, 3Pfizer Pharma GmbH, Berlin, Germany
OBJECTIVES

:
Despite effective vaccines, tick borne encephalitis (TBE) remains an important , sometimes even lethal disease in humans in Germany. The Standing Committee on Vaccination in Germany recommends vaccination of individuals living or working in and traveling to endemic areas, as defined by the Robert Koch-Institute. These "risk areas" are defined if observed human TBE incidences are significantly higher than expected (case-based risk definition) while the number of autochthonous TBE-cases outside of risk areas markedly increase in recent years. Aim of our study is to predict areas in Germany with an increased density of TBEV-infected ticks and thus, to estimate human TBE-cases (exposure-based risk definition).

METHODS

:
We developed a dynamic model to describe the vector–virus–host interaction and its spatio-temporal change. The population is divided into five different health states of “hosts” (i.e. susceptible, infected, recovered, vaccinated humans) and “vectors” (i.e. infected ticks). For each health state, a differential equation predicting the dynamics was implemented. Further, model parameters such as virus transmission or vaccination rates are defined.

RESULTS

:
The model fairly accurately simulates the annual clinical TBE cases in Austria (period 1979-2017) and in Germany (period 2001-2017). Estimated Root-mean-square errors (RMSE) were 68.3 for the Austrian time series and 83.2 for the German time series. Thus far, the model is only forced by the vaccination rate, human demographics, and tick density, the latter depending on mean temperature of the preceding year and the fructification index of European beeches two years prior.

CONCLUSIONS

:
The model currently predicts the smoothened time series of annual human TBE cases on a national level. It is a first step towards developing a novel, exposure-based risk map for human TBE infections. The dynamic model approach also allows testing of hypotheses, as well as forecasting. To further develop the model, several projects (e.g. German-wide tick monitoring) have been initiated.