The risk of arbovirus transmission increases as mosquito infection rates increase: Using simple models to evaluate this assumption

The “infection rate” is an estimate of the arbovirus infection prevalence in a mosquito population. It is used in surveillance to assess risk of virus transmission to humans and domestic animals. It is assumed that risk increases as “infection rate” increases. However, not all infected mosquitoes are involved in transmission thus here we assessed that assumption by investigating (a) how the infection prevalence in the mosquito population relates to the proportion of those mosquitoes that are capable of virus transmission (infectious), and (b) how the infection prevalence relates to “infection rates” obtained after sampling, pooling, and testing mosquitoes for virus. A model was used to approximate the proportion of infected mosquitoes in a cohort that become infectious as a function of temperature, mosquito and virus species. Results suggested that the relationship between proportions of infected and infectious mosquitoes is not linear, and that similar “infection rates” estimated at different times, under different temperatures, or for different mosquito-virus systems may not indicate the same level of risk. Numerical simulations were used to study the “infection rate” as an estimate of the prevalence of infection in the mosquito population. Numerical simulations suggested that “infection rates” usually underestimate prevalence and that they only show true changes in prevalence under certain sampling and testing conditions. “Infection rates” appear not directly related to mosquito infectiousness and risk of transmission, and those rates must be interpreted together with other biological and environmental factors in order to provide a better assessment of the risk of transmission.