Cost Effectiveness of Use of Urine Based Lateral-Flow Lipoarabinomannan Test in Diagnosis of Tuberculosis in HIV-positive Patients in Namibia

Abstract

Background: Namibia is only notifying 62% of the estimated TB cases. Xpert MTB/RIF (Xpert) is considered the gold standard for tuberculosis (TB) diagnosis. The urine lipoarabinomannan (LF-LAM) assay has been recommended to provide rapid TB diagnosis in HIV positive patients to improve case finding. Objectives: To evaluate the cost-effectiveness of LF-LAM when used in combination with Xpert compared to Xpert only for TB diagnosis in an HIV-positive population in Namibia. Method: This was a cost effectiveness study from a health system perspective. A model based on the decision tree was utilized to compare the two strategies for TB diagnosis. A cohort of 1000 patients was assumed. The primary outcomes were the costs and the DALYs averted for each algorithm. The incremental cost-effectiveness ratio (ICER) was determined. Results: The combination of Xpert and LF-LAM compared to Xpert testing alone was very highly cost-effective at ICER of $601 per DALY averted. The willingness to pay (WTP) threshold was the annual GDP per capita ($9,383.20) of Namibia. The cost-effectiveness was most sensitive to the probability of treatment success on first line TB regimen, probability of a positive test for both algorithms as well as the cost of ART but still remained below the WTP threshold. Conclusion: The addition of urine LF-LAM to Xpert test for TB diagnosis in HIV infected individuals is very highly cost effective when compared to Xpert alone. Even after variation of the parameters the ICER remains within WTP in the one-way sensitivity analysis.