Measurement error robustness of a closed-loop minimal sampling method for HIV therapy switching

Area of Science:

• Computational biology
• Pharmacometrics
• Infectious disease modeling

Background:

• Optimizing antiretroviral therapy (ART) in HIV management requires timely adjustments based on viral load (VL) dynamics.
• Switching therapy at or near the VL minimum maximizes therapeutic benefit, as benefit decreases logarithmically with higher VL at the switch time.
• Accurate VL monitoring is crucial, but measurement errors and sampling frequency can impact treatment decisions.

Purpose of the Study:

• To evaluate the robustness of a closed-loop treatment scheduling algorithm against realistic HIV viral load measurement errors.
• To determine the algorithm's ability to accurately detect viral load minima using a reduced number of samples.
• To assess the impact of measurement error on the algorithm's performance in optimizing HIV therapy switch timing.

Main Methods:

• The closed-loop scheduling algorithm was tested against simulated patient data derived from data-driven models.
• Simulated data incorporated experimentally validated log-normal noise to mimic realistic viral load measurement errors.
• Algorithm performance was quantified using metrics such as samples saved, risk reduction, and deviation from optimal switching time.

Main Results:

• The closed-loop algorithm demonstrated robustness when subjected to realistic viral load measurement errors.
• The method achieved significant sample reduction compared to fixed-rate sampling strategies.
• Performance metrics indicated effective risk reduction and near-optimal switching time detection despite noise.

Conclusions:

• The validated closed-loop scheduling algorithm is a robust tool for optimizing HIV treatment strategies.
• The algorithm facilitates accurate detection of viral load minima, enabling timely therapy switches even with measurement error.
• This approach holds potential for improving patient outcomes and reducing healthcare costs through optimized sampling and treatment adjustments.