User Guide

TBsim is a comprehensive tuberculosis simulation framework built on the Starsim platform, designed for modeling TB transmission, natural history, and intervention impacts in realistic populations.

Core Framework

TBsim leverages the Starsim individual-based modeling framework with the following key capabilities:

Individual-Based Modeling (IBM)

Each person in the simulation is represented as an individual agent with unique characteristics, health states, and behaviors.

Network-Based Transmission

Social connections between individuals are modeled as networks, enabling realistic transmission dynamics through household and community structures.

Modular Architecture

Components can be mixed and matched to create custom simulation scenarios with flexible parameterization.

TB Natural History and Disease States

TBsim implements a comprehensive TB natural history model with the following disease states:

Core Disease States

  • Susceptible (NONE): No TB infection

  • Latent TB (LATENT_SLOW/FAST): Two progression rates for latent infection

  • Active Pre-symptomatic (ACTIVE_PRESYMP): Early active disease before symptoms

  • Active Symptomatic Smear+ (ACTIVE_SMPOS): Most infectious form of TB

  • Active Symptomatic Smear- (ACTIVE_SMNEG): Moderately infectious TB

  • Extra-pulmonary TB (ACTIVE_EXPTB): TB outside the lungs

  • Protected (PROTECTED): Protected state from vaccines or preventive therapy

  • TB Death (DEAD): TB-related mortality

Transmission Dynamics

  • Airborne transmission through social networks

  • Smear-based infectiousness (smear+ > smear- > extra-pulmonary)

  • Age-dependent transmission rates

  • Individual transmission heterogeneity

  • Household and community network structures

Diagnostic Capabilities

TBsim includes comprehensive diagnostic modeling with multiple test types:

Available Diagnostic Tests

  • Smear Microscopy: Traditional sputum smear testing

  • Xpert MTB/RIF: Molecular testing for TB and rifampin resistance

  • MDR Testing: Multi-drug resistance detection

  • FujiLAM: Urine-based testing for HIV+ individuals

  • CAD CXR: Computer-aided chest X-ray interpretation

  • Oral Swab: Non-sputum based testing

Diagnostic Features

  • Age and HIV-specific sensitivity/specificity parameters

  • TB state-specific test performance

  • False negative handling with care-seeking multipliers

  • Enhanced diagnostic interventions with comprehensive result tracking

Treatment and Interventions

Drug Treatment

  • DOTS: Directly Observed Treatment, Short-course implementation

  • Latent Treatment: TB Preventive Therapy (TPT) with multiple regimens

  • Treatment Success/Failure: Configurable treatment outcomes

  • Treatment Relapse: Post-treatment disease recurrence modeling

Preventive Interventions

  • BCG Vaccination: Comprehensive BCG protection with age targeting

  • TB Preventive Therapy (TPT): Household-based and individual TPT

  • Vaccine Campaigns: Mass vaccination strategies

Health Seeking Behavior

  • Configurable care-seeking rates

  • Care-seeking multipliers for different populations

  • Re-seeking behavior after treatment failure

  • Age-dependent health seeking patterns

Disease Connectors and Co-infections

TBsim includes sophisticated connectors to link TB with other diseases:

TB-HIV Co-infection

  • HIV State Integration: Uses HIV disease states (ACUTE, LATENT, AIDS) to modify TB progression

  • CD4-dependent Effects: TB progression varies with HIV disease stage

  • ART Effects: Antiretroviral therapy impacts on TB natural history

  • Risk Multipliers: HIV increases TB activation risk (ACUTE: 1.22x, LATENT: 1.90x, AIDS: 3.0x)

TB-Malnutrition Integration

  • Comprehensive Malnutrition Modeling: BMI-based nutritional status effects

  • Nutrition-TB Connector: Nutritional status impacts TB progression and treatment

  • Supplementation Effects: Nutritional intervention modeling

Networks and Social Structure

Network Types

  • Household Networks: Family and household connections for transmission

  • Community Networks: Broader social interactions

  • Trial-based Networks: Framework for intervention trials

  • Contact Networks: Detailed contact tracing capabilities

Transmission Networks

  • Household-based transmission modeling

  • Age-stratified contact patterns

  • Network-based intervention targeting

Analysis and Visualization Tools

TBsim provides comprehensive analysis capabilities through the DWT (Dwell Time) Analyzer system:

Dwell Time Analysis

  • Real-time Tracking: Records time spent in each TB state during simulation

  • State Transition Analysis: Detailed progression pathway tracking

  • Statistical Analysis: Comprehensive dwell time statistics

Visualization Capabilities

  • Sankey Diagrams: Interactive state transition flows

  • Network Graphs: Transmission network visualization

  • Interactive Plots: Plotly-based interactive visualizations

  • Kaplan-Meier Curves: Survival analysis for TB progression

  • Histogram Analysis: Dwell time distribution analysis

  • Reinfection Analysis: Multiple infection episode tracking

Reporting Features

  • Treatment Events: Comprehensive treatment outcome tracking

  • Diagnostic Events: Test result and diagnostic pathway analysis

  • Mortality Events: TB-related death tracking

  • Age-stratified Reports: Age-specific analysis capabilities

  • TB-HIV Reports: Co-infection specific reporting

  • Household Analysis: Household-level intervention impact assessment

Advanced Features

Multi-Scenario Analysis

  • Run multiple parameter combinations simultaneously

  • Comparative analysis across intervention scenarios

Time-varying Parameters

  • Dynamic parameter adjustment during simulation

  • Seasonal and temporal variation modeling

Trial Framework Support

  • Household-based intervention trials

  • Randomized controlled trial simulation capabilities

Calibration and Optimization

  • Automated parameter fitting to observed data

  • Intervention strategy optimization

  • Sensitivity analysis for parameter uncertainty

Running Simulations

Basic simulation workflow:

  1. Setup: Configure TB model, networks, and interventions

  2. Initialize: Create population with demographic data

  3. Run: Execute simulation with real-time analysis

  4. Analyze: Process results with comprehensive visualization tools

  5. Export: Save results for further analysis and reporting

Example Usage

```python import starsim as ss from tbsim import TB, DwtAnalyzer

# Create simulation with TB model and analyzer sim = ss.Sim(

diseases=[TB()], networks=ss.HouseholdNet(), analyzers=[DwtAnalyzer(scenario_name=”Baseline”)]

) sim.run()

# Generate comprehensive analysis plots analyzer = sim.analyzers[0] analyzer.sankey_agents() analyzer.network_graph() analyzer.kaplan_meier() ```

For specific examples and tutorials, see the Examples and Tutorials sections.