Modelling the transmission dynamics of Ross River virus in Southwestern Australia
Institute of Mathematics and its Application
Computing, Health and Science
Engineering and Mathematics
During the 1995–1996 Australian financial year, over 1300 notifications of Ross River (RR) virus disease were notified in humans from Southwestern Australia. Due to the mild symptoms of the disease, it is difficult to diagnose and subclinical infections are common. However, these subclinical infections do give rise to immunity. For planning and control, it is important for public health authorities to estimate the true number of people who have contracted the disease and to assess the impact of key epidemiological parameters. A mathematical model was developed to describe the transmission of RR virus between its hosts (humans and kangaroos) and its vectors (mosquitoes). For this model, the threshold conditions and relative removal rates were calculated and interpreted. Finally, a computer program was written to simulate the model in order to estimate the total number, both clinical and sub clinical human infections given known and hypothetical epidemiological parameter values. Within this simulation sensitivity of the results to changes in the parameters were examined. The analysis of the threshold conditions conformed well to established principles of arboviral transmission and control. It was observed that conditions which can prevent an outbreak occurring include reducing the number of susceptibles in host and vector populations, reducing the infection rates between hosts and vectors and shortening the duration of viraemia. Results on the sensitivity analysis showed that some parameters such as the extrinsic incubation period, mosquito mortality rate in winter and the proportion of Western Grey Kangaroos in the marsupial population have little effect on human incidence. However, the transmission rate between hosts and vectors, vector‐mortality rate in summer and the proportion of infectious vectors among infected vectors have pronounced effects. The simulation results on the ratio of clinical to subclinical human infections predicted a minimum ratio of 1:2 and a maximum ratio of 1:65, which is consistent with data obtained during previous sero‐epidemiological studies.