/MATUMAINI

EMOD Model of co-morbidity of HIV and CMDs

Primary LanguagePython

MATUMAINI - Modeling the co-morbidity of HIV and CMD

  • How does behavioral health impact HIV and HIV treatment outcomes?
  • Common Mental Disorders (CMD) - depression and anxiety
    • can disrupt management of HIV - reducing engagement with care, testing, adherence, retention
    • can lead to risky behaviors which increase exposure risk
  • HIV
    • positive diagnosis can increase risk of CMD through feelings of hopelessness, isolation, stigmatization, shame
    • successful treatment of HIV can reduce risk of CMD

Project Outline

  • Next generation of depression treatment implementation
  • Implement model of CMD within EMOD
    • Add CMD Individual Property (IP) to EMOD which tracks individual CMD status
      • Add CMD IP to the Demography File: CMDStatus
      • Add CMD IP to the config file - report on IP
      • Check: run calibrated model
      • Revisit demographics and aging and how it interacts with recovery, relapse, and treatment history
        • New mathematical model allows for already-depressed individuals to enter the next age bracket with memory of depression history
        • Check: run calibrated model
    • Integrate CMD into model with calibration - Proof of concept
      • Implement simple SIS model of Depression
      • Try calibrating with proof of concept
        • Add CMD IP to dtk_post_process
        • Add CMD prevalence calibration target to the ingest form
        • Add CMD as a variable to be calibrated
    • Develop R analysis tools for postprocessing
    • Develop more detailed CMD model
      • Add age and gender-related incidence modifiers
      • Add different relapse rates for Treated and Untreated individuals
      • Add differentiation between treatment and non-treatment
      • Track individual treatment history, such that individuals who have received treatment are more likely to seek treatment again, less likely to have relapse
  • Implement interactions from CMD to HIV
    • Increased risky behavior among those depressed
      • Transition depressed individuals to MEDIUM risk
      • Transition depressed individuals back when they recover, at appropriate rates
    • Delays to testing - interrupt HCT Testing Loop
      • Add a check on HCTTestingLoop1 - HCTTestingLoopRapidTest
    • ART adherence - differential dropout rates
    • ART VLS outcomes
      • Added effective/noneffective ART choice based on CMD status
      • Added ARTMortality table
      • Added transition from noneffective to effective ART following CMD recovery
    • Delays to initiating ART
  • Implement interactions from HIV to CMD
    • Elevated depression incidence among HIV positive - CoitalActRiskFactors
    • Adjust Risk Group to MEDIUM for depressed individuals
    • Higher depression upon receiving diagnosis
    • Increase depression recovery rate upon reaching VLS
    • Linkage to treatment upon receiving depression diagnosis
  • Future work: other preventative measures
    • PrEP
    • Condom use
    • VMMC
  • Future work: Depression model demographics and dynamics
    • Add CMD-related excess mortality
  • Re-calibrate full model with CMD

Major Updates

  • 2024-09-07 - Generated results for first manuscript on croi branch
  • 2024-03-03 - Presented results from croi branch at CROI 2024
  • 2023-10-31 - Presented results from nimh branch at Global Mental Health Research without Borders Conference
    • Need to revisit treatment and diagnosis scenarios:
      • link to depression care -> screening -> diagnosis -> treatment response
      • floated the possibility of implementing depression treatment distribution in the same way as PrEP distribution