Skip to content

machado-lab/MHASPREAD-model

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

34 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

MHASpread: A Metapopulation Framework for Animal Disease Spread and Control

SEIR_model Stochastic Multi-species Spatial_Transmission Control_Actions

Overview

MHASpread is a comprehensive stochastic, multiscale transmission model designed to simulate epidemic trajectories of foot-and-mouth disease (FMD) and other multi-host pathogens across livestock farming networks. The framework integrates within-farm disease dynamics, spatial transmission processes, and realistic disease control interventions to enable rigorous evaluation of outbreak response strategies.

The model operates at two coupled scales:

  • Within-farm scale: Individual-based SEIR dynamics for multiple host species
  • Metapopulation scale: Network-level transmission via spatial kernels and animal movements

MHASpread is particularly suited for:

  • Epidemiological research: Understanding multi-host disease dynamics and species-specific transmission
  • Policy evaluation: Assessing the effectiveness and cost-efficiency of control strategies
  • Emergency preparedness: Simulating outbreak scenarios and informing contingency planning
  • International capacity building: Supporting training and knowledge transfer in disease surveillance and control

Scientific Background

Foot-and-mouth disease (FMD) is a highly contagious viral disease of cloven-hoofed livestock with significant economic and trade implications. FMD demonstrates complex transmission dynamics due to:

  • Multi-host nature: Differential susceptibility and transmissibility across cattle, swine, and small ruminants
  • Environmental persistence: Heterogeneous transmission via direct contact, aerosol, and fomites
  • Spatial heterogeneity: Disease spread constrained by distance but facilitated by animal movements
  • Species-specific traits: Varying latent periods, infectious periods, and recovery rates

MHASpread explicitly models these complexities to provide mechanistic insights into outbreak progression under diverse epidemiological and management scenarios.


Model Description

MHASpread is built on a stochastic, SEIR-based framework with distinct compartments for each host species:

Compartment Definition
S Susceptible: animals not infected and able to acquire infection
E Exposed: infected but not yet infectious (latent period)
I Infectious: infected animals capable of transmitting infection
R Recovered: animals recovered with temporary or permanent immunity
V Vaccinated: animals with vaccine-induced immunity

Within-Farm Dynamics

Disease progression within farms follows species-specific transmission probabilities (β) and disease duration parameters:

  • Transmission coefficient (β): Determined by both infected and susceptible species, reflecting differential transmissibility (e.g., swine are more efficient FMD spreaders than cattle)
  • Latent period (σ): Average 2–5 days, species-dependent
  • Infectious period (γ): Average 3–6 days, species-dependent
  • Vital dynamics: Births and deaths incorporated where data available

Spatial Transmission

Local spread between farms is modeled using an exponential transmission kernel:

$$P_E(t) = 1 - \prod_i \left(1 - \frac{I_i(t)}{N_i} \phi e^{-\alpha d_{ij}}\right)$$

where:

  • $d_{ij}$ = distance between farms (maximum 40 km)
  • $\phi$ = baseline transmission probability (0.044)
  • $\alpha$ = kernel decay parameter (0.6)
  • $I_i(t)/N_i$ = infection prevalence at farm $i$

This formulation captures the empirically observed pattern of decreasing transmission risk with distance, bounded by documented maximum dispersal distances.

Disease Detection

Infected farms are detected through active surveillance using a hypergeometric sampling process that:

  1. Inspects a fraction of farms under surveillance (≥1 farm)
  2. Identifies infected farms accounting for finite surveillance population
  3. Incorporates diagnostic imperfection (sensitivity $s$)
  4. Tracks traceback-identified farms through contact tracing

Control Interventions

MHASpread implements four primary control strategies:

Control Description Implementation
Depopulation Culling of infected farms Priority by herd size; daily capacity limits
Emergency Vaccination Ring vaccination of bovine herds Infected + buffer zones; 15-day lag; daily capacity limits
Movement Standstill Restriction on animal movements 30-day duration across infected, buffer, and surveillance zones
Contact Tracing Identification of linked farms 30-day traceback window; farms under enhanced surveillance

Control Zones

Three nested zones enforce differential surveillance and intervention intensity:

  • Infected zone (3 km): Focus of depopulation and immediate control
  • Buffer zone (7 km): Secondary vaccination prioritization
  • Surveillance zone (15 km): Active detection and monitoring

Key Features

Multi-host dynamics: Explicit modeling of cattle, swine, and small ruminants
Species-specific transmission: Parameterized from empirical literature
Stochastic processes: Incorporates uncertainty in transmission, detection, and control efficacy
Spatial heterogeneity: Distance-dependent transmission and realistic farm networks
Flexible control actions: Customizable intervention timing, capacity, and extent
Diagnostic imperfection: Reflects real-world detection limitations
Vital dynamics: Optional incorporation of births and deaths
Scenario analysis: Supports evaluation of multiple strategies


Applications

MHASpread has been applied to:

  • FMD preparedness in South America: Evaluating control strategies for Brazil (2024) and Bolivia (2023)
  • Cost-effectiveness analysis: Integrating epidemiological and economic models to inform policy
  • Surveillance optimization: Assessing detection capacity and traceback effectiveness
  • International training: Building epidemiological modeling expertise in partner countries

Events and Training

The MHASpread framework has been instrumental in international workshops and capacity-building initiatives aimed at strengthening disease surveillance and control expertise among veterinary and public health authorities in Latin America.

Chile FMD Workshop 2024

workshop_aftosa_chile_2024
Training program for Chilean national authorities on FMD simulation modeling, risk-based surveillance design, and emergency response preparedness. Participants gained hands-on experience with MHASpread to evaluate control strategies tailored to Chilean agricultural contexts.

PANAFTOSA Workshop Rio 2023

PANAFTOSA-Workshop-Rio2023
Regional capacity-building initiative hosted by PAHO's Pan American Animal Health Organization, bringing together epidemiologists and veterinary officials from across the Americas. Focus on metapopulation modeling, uncertainty quantification, and evidence-based contingency planning for transboundary animal diseases.

PAHO MHASpread Workshop Repository

MHASpread_workshop_PAHO
Comprehensive open-access workshop materials including tutorials, datasets, and worked examples. Provides reproducible workflows for disease simulation, sensitivity analysis, and policy evaluation using MHASpread-based methodologies.


Relationship to Published Work

MHASpread development and application are documented in peer-reviewed publications:


Limitations

  • Spatial scope: Model designed for farm-level networks; may require adaptation for larger-scale systems
  • Data requirements: Accurate population sizes, movement records, and species composition needed
  • Control realism: Model assumes adherence to protocol; deviations may affect outcomes
  • Single-pathogen focus: Not designed for multi-pathogen coinfection scenarios

Citation

If you use MHASpread in your research, please cite:

Cespedes Cardenas, N., & Machado, G. (2024). Modeling foot-and-mouth disease dissemination in Brazil and evaluating the effectiveness of control measures. Frontiers in Veterinary Science, 11, 1468864. https://doi.org/10.3389/fvets.2024.1468864


License

This repository contains documentation and educational materials for MHASpread. The proprietary model source code is provided under a separate commercial or institutional license agreement. See LICENSE.md for details.


Acknowledgments

MHASpread development is supported by:

  • FUNDESA RS (Fundação de Desenvolvimento para Excelência em Ciência e Tecnologia do Rio Grande do Sul)
  • NC State University - Department of Population Health and Pathobiology
  • LUMAC (Laboratório de Unidades Multidisciplinares de Apoio Científico) — Universidade Federal de Santa Maria

Developers

🖥️ Nicolas Cespedes Cardenas ORCID
Machado Lab, College of Veterinary Medicine
NC State University

🖥️ LUMAC Team
Universidade Federal de Santa Maria, Brazil


Documentation

For detailed technical documentation, visit our comprehensive documentation site or explore the sections below:

About

Spread disease model in R flavored with python

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors