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

• 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

• 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