Adaptive modeling of HIV-TB coinfection dynamics and intervention optimization

Background: The syndemic of HIV and tuberculosis (TB) co-infection remains a critical global health challenge, particularly in resource-limited settings where conventional epidemiological models fail to capture the complex evolutionary dynamics between pathogens, hosts, and interventions. Current approaches lack adaptive mechanisms to account for temporal changes in transmission parameters and quality of life (QoL) impacts, creating an urgent need for innovative modeling frameworks. Methods: This study focuses on the Khyber Pakhtunkhwa province of Pakistan, a region with moderate HIV prevalence and high TB incidence. The analysis used de-identified clinical and demographic data (N = 592) collected from tertiary hospitals in KP between 2021 and 2023. We developed a novel hybrid modeling approach integrating empirical clinical data with evolutionary computation through three synergistic components: (1) evolutionary-optimized demographic sampling (), (2) time-varying compartmental modeling with adaptive parameters (), and (3) multi-objective intervention optimization. The framework was validated through a four-pillar approach incorporating statistical metrics, evolutionary cross-validation, clinical evaluation, and policy impact assessment. Key results: Our analysis revealed three critical findings: First, transmission parameters exhibited distinct temporal patterns, with TB showing saturating growth () while HIV declined gradually. Second, gender-specific exposure dynamics were identified, with males having significantly higher transmission risk (,). Third, the targeted treatment strategy emerged as optimally cost-effective (ICER = $2,300/quality-adjusted life years (QALY), 95% CI: 1,850–2,750), reducing by 54% while maintaining high feasibility (0.91). Significance: This study provides the first comprehensive framework that simultaneously addresses pathogen evolution, host dynamics, and intervention optimization in HIV-TB co-infection. The findings offer actionable insights for public health policy, particularly in balancing cost-effectiveness with implementation feasibility. Our evolutionary-optimized approach establishes a new paradigm for modeling complex disease systems, with potential applications extending beyond HIV-TB to other interacting epidemics.