Developing dengue early detection model for enhancing dengue mitigation strategies: a case of Bhopal city

Abstract

Dengue, a vector-borne disease transmitted by Aedes mosquitoes, presents a significant global health threat, with escalating annual case numbers. In India alone, annual incidences of dengue rose from 12,561 cases and 80 deaths in 2008 to 2,33,251 cases and 303 deaths in 2022. The disease's transmission dynamics are intricately linked to meteorological and physical factors, further worsened by climate change, which expands suitable habitats for disease vectors, potentially exposing millions to a previously unrecognized threat. Despite numerous studies exploring the relationship between dengue cases and environmental factors, few have developed predictive models, especially at the city scale, with limited implementation in vector surveillance, particularly in India. This thesis addresses this gap by focusing on improving early detection strategies, particularly in Bhopal Municipal Corporation (BMC), where environmental conditions favour dengue breeding. Leveraging RS&GIS, the study aims to develop a Dengue Early Detection (DED) Model to improve dengue mitigation strategies. Through bibliometric analysis, key spatial-temporal variables influencing dengue transmission were identified such as Temperature, Rainfall, Humidity, NDVI, Landcover, Topological Wetness Index, and Population Distribution. These variables were processed using RS&GIS and converted into monthly ward-wise data, to be incorporated into the model. The relationships between independent and dependent variables were also studied highlighting the existing spatial-temporal biases within the data. The DED Model was developed using the Sequential Regression Model with Quasi-Poisson Generalized Linear Model showing the ability to predict 40% of the variance of dengue cases (Adjusted Pseudo-R2 = 0.407, RMSE Train = 2.769 and RMSE Test = 1.810). For contextualizing the model at the local level, a high-risk ward within BMC i.e., ward 65 was selected and local factors influencing dengue transmission were mapped such as lack of street maintenance, solid waste management, open space management, water storage practices, and inadequate drainage infrastructure. Further stakeholder interviews were conducted to gain insights about ongoing dengue control efforts and challenges faced by communities. This led to tailored interventions aimed at controlling dengue vector within the ward, encompassing technological, ecological, and community engagement approaches, emphasizing the importance of multifaceted strategies. These included the development of a DED application for sharing information and engaging the community in monitoring breeding sites at the city level, and awareness campaigns on important practices and relevant information throughout the city. Also, infrastructure improvements such as bio-control gardens and bioswales integrated with biological measures, which would involve the community in plantation drives of natural mosquito repellent plants to manage vector population in the ward. This study underscores the effectiveness of RS&GIS technologies in strengthening dengue surveillance and aiding in control measures, allowing timely, targeted, and tailored interventions while advocating for community-driven initiatives to embrace a comprehensive approach. The study also emphasizes the need for continuous monitoring and adaptation of strategies to address the evolving dynamics of dengue transmission for future research and policy implementations aimed at combating vector-borne diseases.