Objectives: The cost-effectiveness of screening adult patients for pulmonary tuberculosis is not clear. As such, this study aims to identify the cost-effectiveness between the Xpert MTB/RIF assay and the sputum acid-fast bacilli (AFB) smear. Multi-outcomes were correct diagnosis, time to achieve correct diagnosis, and gain in quality-adjusted life-years (QALYs).

Methods: A decision tree model was constructed to reveal a possible clinical pathway of tuberculosis diagnosis. The researchers used a clinical study to establish the probability of all clinical pathways for input into this model. The sample size was calculated following the correct diagnosis. Participants were randomly divided into 2 groups. A structural questionnaire and the Thai version of quality of life (EQ-5D-5L) were used for interviewing.

Results: The results showed that the time to achieve the correct diagnosis for the group using Xpert MTB/RIF was shorter than that for the group using the sputum AFB smear. Both the correct diagnosis and QALYs of the base case analysis presented the Xpert MTB/RIF method as dominant. A Monte Carlo model, which analyzed the Xpert MTB/RIF method, revealed that the average number of patients who were correctly diagnosed was 673, the QALYs were 945.85 years, and the total cost was $143 110.64. For the sputum AFB smear method, the average number who received a correct diagnosis was 592, the QALYs were 940.40 years, and the total cost was $196 666.84. Probabilistic and one-way sensitivity analysis confirmed that the Xpert MTB/RIF remained dominant.

Conclusions: These results provide useful information for the National Strategic Plan to screen all adult patients for pulmonary tuberculosis.