Arbovirus models to provide management tools for mosquito control and disease prevention in the Northern Territory, Australia

Abstract

Ross River virus (RRV) causes the most common human arbovirus disease in Australia. Although the disease is nonfatal, the associated arthritis and postinfection fatigue can be debilitating for many months, impacting on workforce participation. We sought to create an early-warning system to notify of approaching RRV disease outbreak conditions for major townships in the Northern Territory. By applying a logistic regression model to meteorologic factors, including rainfall, a post-estimation analysis of sensitivity and specificity can create rainfall cut-points. These rainfall cut-points indicate the rainfall level above which previous epidemic conditions have occurred. Furthermore, rainfall cut-points indirectly adjust for vertebrate host data from the agile wallaby (Macropus agilis) as the life cycle of the agile wallaby is intricately meshed with the wet season. Once generated, cut-points can thus be used prospectively to allow timely implementation of larval survey and control measures and public health warnings to preemptively reduce RRV disease incidence. Cut-points are location specific and have the capacity to replace previously used models, which require data management and input, and rarely provide timely notification for vector control requirements and public health warnings. These methods can be adapted for use elsewhere.