< Back to previous page


Incorporating behavior in infectious disease models for economic evaluation

Book - Dissertation

Epidemiological and economic models are used to simulate the impact of interventions on the spread of infectious diseases. Since behavior is an important factor in this, behavioral change models have been developed. Often, it is assumed that a rational individual will exhibit “free-rider” behavior: relying on herd immunity, they will increasingly refuse vaccination as vaccination coverage increases. This hypothesis originates in game theory. A systematic review of 178 papers showed an increasing trend in published behavioral change models. However, only 15% of these used observed data for parameterization or validation. This is why we studied the relative importance of six vaccine-specific attributes relevant to decisions about vaccination. We conducted a discrete choice-experiment in Flanders, Belgium, South-Africa, France, the Netherlands and the UK. Effectiveness and accessibility of the vaccine were found to be most important. Contrary to what game-theoretical models assume, social norms dominated free-rider behavior. Using these results, we simulated the dynamic coverage of vaccination against measles in Flanders – parallel to a dynamic transmission model. We examined the impact of different scenarios regarding behavior on vaccine utility, vaccine coverage and disease dynamics. Exogeneous shocks, such as a “vaccine scare” or suspension of the vaccination program, caused repeated outbreaks. In the context of an economic analysis on workplace vaccination, we simulated the spread of influenza in Belgium, using a compartmental transmission model. We accounted for both age-specific social contact behavior and reduced contact behavior of symptomatic individuals. Workplace vaccination – funded by the employer – was usually cost-saving, and, furthermore, reduced the burden of disease of influenza, both in the workplace and in the general population. Vaccine-induced herd immunity should thus be regarded as something positive. Indeed, we did not observe free-rider behavior in any of the study-populations. On the contrary, social norms increased the perceived value of vaccines. Communication should thus emphasize that vaccination is still the norm. Another important factor in vaccination decisions was accessibility. Regarding the latter, there is still much room for improvement for vaccines that target the adult population. In particular, workplace vaccination against influenza can be an alternative to time-consuming visits to both the doctor’s office and the pharmacy. Finally, given the non-linear relationship between the uptake of prevention measures and disease dynamics, epidemiological and economic models should take behavior change into account. However, to develop behavior change models that are not purely theoretical – as indeed purely game-theoretical models are outdated - more data collection is necessary.
Edition: 99
Volume: 99
Number of pages: 290
Publication year:2020
Embargoed until:09/11/2022