Cost-effectiveness of laboratory monitoring for management of HIV treatment in sub-Saharan Africa

OBJECTIVE: To compare the cost-effectiveness of three different strategies for long-term monitoring of antiretroviral therapy (ART) failure and regimen switching in sub-Saharan Africa: a symptom-based approach, or monitoring of either CD4 cell counts or plasma viral load (pVL). DESIGN: Markov model. SETTING AND PARTICIPANTS: Hypothetical HIV-infected adult population who began first-line ART and subsequently had up to 6 years of follow-up. MAIN OUTCOME MEASURES: Total cost, life expectancy and incremental cost-effectiveness ratio (ICER). RESULTS: A symptom-based approach yielded a life expectancy of 64.0 months at a total cost of US$ 4028 per person. All laboratory-based strategies, at testing intervals of 6 or 12 months, were cost-saving and improved life expectancy, compared with a symptom-based approach. The life-expectancy gain was larger for pVL than for CD4 strategies at 6-monthly (2.3 and 0.9 months, respectively) and 12-monthly testing (2.0 and 0.8 months, respectively). Cost-savings of 6-monthly pVL or CD4 testing were similar (US$ 630 and 621, respectively), whereas 12-monthly CD4 cell counts were more cost-saving than 12-monthly pVL (US$ 1132 and 880, respectively). Testing every 12 months - rather than every 6 months - decreased the ICER by 102% for CD4 cell count and 67% for pVL. These findings were robust to a wide range of deterministic sensitivity analyses, but were sensitive to the specificity and costs of diagnostic tests. CONCLUSION: Additional diagnostic costs are balanced by cost-savings from avoiding unnecessary switching due to misdiagnosis of ART failure. Routine pVL monitoring may be preferred as a replacement for CD4 cell counts because of its additional public-health advantages in preventing drug-resistance, supporting adherence and reducing HIV transmission.