Spatial transferability of an agent-based model to simulate Taenia solium control interventions

Area of Science:

• Epidemiology
• Mathematical Modeling
• Parasitology

Background:

• Taenia solium (T. solium) is a zoonotic parasite causing epilepsy and economic losses.
• Mathematical models aid in predicting the impact of T. solium control interventions.
• Model calibration is crucial for credibility but can limit geographic transferability.

Purpose of the Study:

• To assess if a non-local model calibration approach enhances the spatial transferability of the CystiAgent model for T. solium transmission.
• To evaluate the model's ability to predict transmission dynamics in new locations without recalibration.

Main Methods:

• Employed a non-local calibration approach using T. solium data from eight villages in Northwest Peru.
• Transferred the calibrated CystiAgent model to 21 additional villages.
• Validated model outputs against 2-year pig serology data from intervention trials.

Main Results:

• Simulated pre-intervention pig seroprevalences were validated in 81% of destination villages.
• Model outputs reproduced validated pig seroincidence in 76% of destination villages post-intervention.
• Demonstrated successful spatial transferability of the CystiAgent model without recalibration.

Conclusions:

• The non-local calibration approach enables robust T. solium transmission modeling across geographically similar regions.
• The model can simulate baseline conditions and intervention outcomes, supporting large-scale regional assessments.
• Facilitates the development of regional models for T. solium transmission and control.