Top Banner
The role of isolation, self-protection, and mosquito abatement in an Agent-Based model of Dengue Fever transmission Fred W. Selck , Nidhi Bouri , Crystal Franco , Tara Kirk Sell , Amesh A. Adalja Dengue fever (DF), a neglected tropical disease spread by mosquitoes has traditionally been described in compartmental models. Agent-based models provide an alternative means of quantifying the spread of dengue and the effect of various interventions. An agent-based model was developed in which humans are shifted between 2 discrete environments: (1) “inside”, in which human-to- mosquito contact is minimized (i.e. through screens, air conditioning, etc.) and (2) “outside”, in which most or all of the human-to-mosquito contact occurs. These environments consist of discrete patches on a two-dimensional lattice and the assignment of individuals to the outside patch is randomly determined. The outside patches have population gradients randomly assigned to reflect variations in mos- quito populations observed in nature. Introduction of dengue occurs in a naive mosquito environment through the introduction of infected individuals into the native human population. Individuals are assigned as either infected or exposed, conditional upon successful transmission when in the outside patch. Agent behavior (AB) can be modified in one of two ways: individuals stay inside when they are infected, and individuals adopt self-protec- tion measures (i.e. repellant application). The latter is conditional upon the prevalence of DF and vector population size – both of which are decision rules applied prior to the model start. Dengue transmission is dependent upon the bite rate (conditional upon mosquito density and human self-protection). Modifications to mosquito population levels (i.e. through mosquito abatement strategies) are exogenously applied to the model. Simulations are run in discrete time until an outbreak reaches equilibrium. The model was created using NetLogo version 4.1.3 (Northwestern University, 2011). The preliminary results suggest that all three strategies (self-protection, abatement, and isolation) are important determinants in DF transmission. After multiple simulations, the reproductive rate for varying mosquito densities can be determined. The effect of various mitigation strategies on this parameter can be tested via this model. This agent-based model can provide important insights into the relationship between individual agent behaviors and DF transmission. In contrast to other modeling strategies, the agent-based approach allows for both determined and stochastic changes in agent (human) behavior which can significantly affect DF transmission dynamics. Contact: [email protected] Background Methods and Materials Results Conclusion Figure 4 Figure 3 Figure 2 Figure 1 Fig 1. Latently infected individuals (pink) are imported into the environment. Individuals move from inside (left) to outside (right) once a day. Mosquito population gradients are indicated in grayscale. Fig 2. After an 8 day incubation period, exposed mosquito patches are infectious (in yellow). Fig 3. Infected cases, both asymptomatic (pink) and symptomatic (red), develop after 14 days. Symptomatic individuals remain indoors for the duration of their infectious period. Fig 4. Individuals adopt mosquito repellent (green) after a defined threshold of infectious cases are reported. 1 2 2 2 2 Johns Hopkins Bloomberg School of Public Health, Center for Biosecurity, Center for Biosecurity 2 1
1

Dengue Modeling Paper - Johns Hopkins Center for Health ...

Jan 05, 2022

Download

Documents

dariahiddleston
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Dengue Modeling Paper - Johns Hopkins Center for Health ...

The role of isolation, self-protection,and mosquito abatement in an Agent-Based

model of Dengue Fever transmissionFred W. Selck , Nidhi Bouri , Crystal Franco , Tara Kirk Sell , Amesh A. Adalja

Dengue fever (DF), a neglected tropical disease spread by mosquitoes has traditionally beendescribed in compartmental models. Agent-based models provide an alternative means of

quantifying the spread of dengue and the e�ect of various interventions.

An agent-based model was developed in which humans are shifted between 2 discrete environments: (1) “inside”, in which human-to-mosquito contact is minimized (i.e. through screens, air conditioning, etc.) and (2) “outside”, in which most or all of the human-to-mosquito contact occurs. These environments consist of discrete patches on a two-dimensional lattice and the assignment of individuals to the outside patch is randomly determined. The outside patches have population gradients randomly assigned to re�ect variations in mos-quito populations observed in nature.

Introduction of dengue occurs in a naive mosquito environment through the introduction of infected individuals into the native human population. Individuals are assigned as either infected or exposed, conditional upon successful transmission when in the outside patch. Agent behavior (AB) can be modi�ed in one of two ways: individuals stay inside when they are infected, and individuals adopt self-protec-tion measures (i.e. repellant application). The latter is conditional upon the prevalence of DF and vector population size – both of which are decision rules applied prior to the model start. Dengue transmission is dependent upon the bite rate (conditional upon mosquito density and human self-protection). Modi�cations to mosquito population levels (i.e. through mosquito abatement strategies) are exogenously applied to the model. Simulations are run in discrete time until an outbreak reaches equilibrium. The model was created using NetLogo version 4.1.3 (Northwestern University, 2011).

The preliminary results suggest that all three strategies (self-protection, abatement, and isolation) are important determinants in DF transmission. After multiple simulations, thereproductive rate for varying mosquito densities can be determined. The e�ect of various mitigation strategies on this parameter can be tested via this model.

This agent-based model can provide important insights into the relationship between individual agent behaviors and DF transmission. In contrast to other modeling strategies, the agent-based approach allows for both determined and stochastic changes in agent (human) behavior which can signi�cantly a�ect DF transmission dynamics.

Contact: [email protected]

Background

Methods and Materials

Results

Conclusion

Figure 4Figure 3Figure 2Figure 1

Fig 1. Latently infected individuals (pink) are imported into the environment. Individuals move from inside (left) to outside (right) once a day. Mosquito population gradients are indicated in grayscale.

Fig 2. After an 8 day incubation period, exposed mosquito patches are infectious (in yellow).

Fig 3. Infected cases, both asymptomatic (pink) and symptomatic (red), develop after 14 days. Symptomatic individuals remain indoors for the duration of their infectious period.

Fig 4. Individuals adopt mosquito repellent (green) after a de�ned threshold of infectious cases are reported.

1 2 2 2 2

Johns Hopkins Bloomberg School of Public Health, Center for Biosecurity, Center for Biosecurity21