ABSTRACT
Patients with HIV experience high mortality and severe morbidity if co-infected with tuberculosis. This is especially true in countries of sub-Saharan Africa like Nigeria with high burden of both diseases where diagnostic capacity for TB in HIV positive patients is poor. Early diagnosis and treatment has been identified as an important strategy to mitigate these effects. A cost and cost-effectiveness analyses of TB diagnostic strategies to reduce this adverse outcomes was carried out. A decision analysis model was developed to estimate the incremental cost, Life years gained, and cost-effectiveness of 3 TB diagnostic algorithms. The base case scenario was an algorithm relying on sputum smear, and chest radiography, which was compared to algorithms featuring newer TB diagnostics (GeneXpert MTB/Rif and MODS). All relevant costs from a National TB programme perspective using the timeframe of the first 6 months of ART was considered. We conducted a one-way sensitivity analysis on parameters in the model. When considering TB diagnosis and treatment costs, the cost per patient was $126.39 for current practice, $76.96 for the algorithm with Xpert test, and $88.62 for the algorithm with MODS. Life years gained was 0.9 for all 3 algorithms. The results showed that both new TB diagnostic methods were more cost effective than base case scenario. The algorithm with Xpert test was cheapest to achieve similar effectiveness compared with the MODS and base case. Sensitivity analyses indicated that cost-effectiveness findings were robust. This study showed that Xpert was cost-effective compared to MODS and base case algorithm. Thus, these findings provide support for ongoing efforts to expand TB diagnostic capacity with GeneXpert MTB/Rif. xii Key Words: HIV, tuberculosis, tuberculosis diagnosis, cost, cost- effectiveness, sub-Saharan Africa, Nigeria.