A Hybrid Agent-Based Model of Urban Dengue Transmission: City Specific Adaptation and Validation in Santa Marta, Colombia
Applied Sciences·
- DOI
- 10.3390/app16168219
- PMID
- —
- PMCID
- —
- OpenAlex
- W7203701065
- Study type
- Journal article
- Publisher
- MDPI AG
- Article type
- journal-article
- Integrity
- current
Why this research matters now
The methodology offers a practical tool for evaluating localized vector control strategies without requiring extensive real-time data collection during active outbreaks.
Structured evidence summary
Research question
How can urban dengue transmission be accurately simulated while maintaining computational efficiency for local calibration and scenario testing?
Study design
This work utilizes a hybrid agent-based computational framework paired with high-performance computing to simulate disease spread. The approach combines individual-level human mobility and infection dynamics with patch-level mosquito population equations. Calibration relied on Bayesian optimization, followed by sensitivity testing and validation against historical outbreak records.
Population and setting
The simulation focuses on an urban environment in Santa Marta, Colombia, representing human inhabitants as mobile entities and mosquito cohorts across defined geographic patches.
Main findings
The calibrated simulation successfully mirrored the overall scale and primary seasonal trajectory of recorded dengue cases. However, it struggled to replicate the prolonged period of minimal incidence toward the end of the transmission season. Simulated interventions targeting mosquito breeding habitats demonstrated substantial potential to reduce overall case counts and interrupt sustained community spread.
Public-health relevance
The methodology offers a practical tool for evaluating localized vector control strategies without requiring extensive real-time data collection during active outbreaks.
Important limitations
The framework did not account for imported infections or precipitation effects on mosquito population growth, which limited its ability to match late-season case patterns. Additionally, the findings derive entirely from computational scenarios rather than empirical field trials.
GIDS interpretation
This publication serves as a technical reference for developers constructing localized computational epidemiology tools. Researchers searching for methodologies that integrate geospatial data and high-performance computing into disease simulation workflows may find this architecture useful for contextual benchmarking.
Related GIDS surveillance
Literature context does not validate, explain, or change a surveillance signal. Exact and contextual relationships are shown separately.
Evidence relationships
This article has 8 auditable classifier relationships to diseases, places, topics, and study design.