Simulating endogenous dynamics of intervention-capacity deployment: Ebola outbreak in Liberia

During the first months, the 2014 outbreak of the Ebola virus (EBOV) in West Africa was characterised by inadequate intervention capacities. In this paper, we investigate (1) the influence of limited but dynamic intervention capacities and their effect on the effective reproduction number, and (2) the effects of proactive versus reactive intervention approaches. We use a transmission model extended with dynamical intervention capacities. Taking into account a bandwidth for potential over- and under-reporting in reported Ebola virus disease cases, the model is used to generate ensembles of plausible scenarios. Next, it is used for testing the effectiveness of more proactive approaches in extending intervention capacities across these scenarios. We show that reactive approaches in extending intervention capacities can lead to continued under-capacity, and, consequently, to an increase of the effective reproduction number and to accelerated EBOV transmission. Proactive approaches, which take deployment delays, doubling times of diseases, and potential under-reporting of the number of cases into account, help in limiting the total number of cases and deaths if the effective reproduction number in isolation is lower than the effective reproduction number outside of isolation. If the effective reproduction number in isolation is higher, proactive intervention policies still outperform reactive intervention policies.