Optimizing HIV Therapy: outcomes, challenges, and opportunities for virological monitoring in resource-limited settings Public Deposited

Downloadable Content

Download PDF
Last Modified
  • March 19, 2019
  • Rutstein, Sarah
    • Affiliation: Gillings School of Global Public Health, Department of Health Policy and Management
  • Despite extraordinary advances in antiretroviral therapy (ART) coverage, only a fraction of the millions of African ART patients have access to routine viral load (VL) monitoring. The goal of this dissertation was to address how to design and implement effective, efficient, and feasible VL monitoring strategies in resource-limited settings. In Aim 1, we studied programmatic and clinical outcomes of dried blood spots (DBS) for VL monitoring, enrolling 1,498 ART patients from five district hospitals in Malawi. Result delivery was a challenge. Providers frequently failed to deliver available results, and nearly half (665/1498) of participants had a clinic visit without receiving results. Nonetheless, 80% of participants received results within 3 months. We observed a lower-than-expected failure rate; only 88 (5.9%) participants had an elevated VL (>5,000 copies/ml) at baseline. Most (92.6%) eligible patients initiated second-line therapy. In Aim 2, we interviewed 17 providers involved in the DBS study. Providers identified a complex set of interconnected barriers and facilitators to VL monitoring. Echoing Aim 1 results, providers described challenges with result delivery and tracking, exposing gaps in data management systems. For many providers, the study was the first time they used an objective marker of ART response to guide clinical management. Provider empowerment emerged as an unexpected facilitator of VL monitoring. In Aim 3, we used data from a Phase IV open-label trial to develop a risk score identifying persons with resistance among those with elevated VLs. Facilitating eventual risk score implementation, we only used parameters likely available to providers in resource-limited settings. We developed three model iterations, increasingly more restrictive in terms of assumptions regarding availability of a patient's laboratory information. The sensitivity for the three models ranged from 10.0%-26.0%, and specificity ranged from 97.4%-99.5%. Our studies identified programs that reliably identified virological failure, are feasible and offer unexpected provider benefits in the resource-limited ART clinical setting, and equip providers with point-of-care algorithms facilitating rapid treatment change for patients with ART resistance. Together, these findings bring us closer towards our longer-term goal of optimizing ART use and improving the quality of ART management and HIV care delivered in resource-limited settings.
Date of publication
Resource type
Rights statement
  • In Copyright
  • Hosseinipour, Mina
  • Wheeler, Stephanie
  • Miller, William
  • Biddle, Andrea K.
  • Weinberger, Morris
  • Doctor of Philosophy
Degree granting institution
  • University of North Carolina at Chapel Hill Graduate School
Graduation year
  • 2015
Place of publication
  • Chapel Hill, NC
  • There are no restrictions to this item.

This work has no parents.