HIV testing criteria to reduce testing volume and increase positivity in Botswana

Area of Science:

• Public Health
• Epidemiology
• Biostatistics

Background:

• Routine human immunodeficiency virus (HIV) testing is crucial for early diagnosis and treatment.
• Optimizing HIV testing strategies is essential to improve efficiency and resource allocation in public health programs.

Purpose of the Study:

• To develop and validate risk scores for identifying undiagnosed HIV infections within a routine testing program.
• To assess the potential of risk scores to reduce the total number of HIV tests performed while maintaining diagnostic yield.

Main Methods:

• Utilized demographic and risk factor data from a large-scale HIV testing program in Botswana.
• Employed multivariate logistic regression to identify factors associated with HIV positivity.
• Developed risk scores based on identified factors and evaluated their performance using prediction and validation datasets.

Main Results:

• Identified 12 significant factors associated with HIV positivity in the second study period.
• Risk score implementation demonstrated potential to reduce testing volume by up to 50% while capturing a high percentage of positive cases (up to 96%).
• Achieved cost savings ranging from 12.1% to 52.3% through reduced testing volume.

Conclusions:

• Risk score-based HIV testing can significantly decrease testing volume and associated costs.
• The strategy may lead to a trade-off between reduced testing volume and potential delays in diagnosing some HIV-positive individuals.
• The clinical and operational implications of this trade-off require careful consideration within Botswana's public health context.