Mathematical Modeling to Support the Interpretation of Spatial Repellent Clinical Trials and Cost-Effectiveness Projections

Mosquito-transmitted diseases such as malaria and dengue are major public health burdens. Interventions that target mosquito vectors are promising for preventing contact between humans and mosquitoes. One such intervention class that is currently being tested in clinical trials is spatial repellents (SRs), which are products that may lower human-mosquito contact by driving mosquitoes away from human-inhabited spaces and/or interfere with the mosquito host-seeking and biting processes. Properly quantifying the mechanism of action of SRs is crucial to projecting the effectiveness of SRs when deployed on a large scale as the effectiveness of SRs is likely to vary geographically due to differences in transmission intensity and climate, among other factors. Three analyses were conducted to quantify SR mechanisms of action and cost-effectiveness. Firstly, a Bayesian model was developed to attribute changes in malaria incidence, malaria prevalence, and human biting rates in a clinical trial in Indonesia to behavioral effects in the \textit{Anopheles} mosquito population induced by SRs. This analysis showed that SRs lowered the mosquito biting rate in treated homes by 25.5% (95% credible interval [CrI]: 4.62-39.7%), reduced mosquito mortality by 7.10% (95% CrI: 16.3% decrease-7.02% increase), and increased the amount of time mosquitoes spent inside human-inhabited spaces by 7.56% (95% CrI: 10.3% decrease-26.0% increase). Secondly, a similar model was developed for a trial of SRs for prevention of arboviruses such as dengue in Peru. This analysis showed that SRs lowered the mosquito biting rate in treated homes by 9.85% (95% CrI: 4.72-15.7%), increased mosquito mortality by 32.6% (95% CrI: 11.7-60.5%), decreased the probability of mosquitoes entering treated homes by 33.4% (95% CrI: 15.1-52.6%) and decreased the exiting rate of mosquitoes from treated homes by 7.81% (95% CrI: 34.0% decrease - 17.4% increase). Thirdly, optimal control theory along with estimates of SR effects from chapter two, a climate-driven model of mosquito abunadance, and an SEI-SEIS model were used to project the cost-effectiveness of SRs when deployed throughout Kenya. We found that cost-effectiveness of SRs depended critically on estimates of SR impact on mosquito mortality, and that SRs are likely to be cost-effective . In conclusion, this dissertation proposes three studies to aid the decision of approval of SRs for use against malaria and dengue and provides a framework for understanding and elaborating on clinical trials of mosquito-borne diseases.