Interventions:
A justification for the selection of the comparators was given. The authors stated that serological tests were faster and simpler to perform than sputum-based methods and other tests, but they were not recommended by any agency.
Effectiveness/benefits:
Little information on the sources of clinical data was provided, but meta-analyses were used if available, as they were for the key model inputs (test accuracies). The design and characteristics of the other sources were not described, making it difficult to fully assess their validity. The authors stated that high variability was found around several model inputs and extensive sensitivity analysis was conducted. DALYs were a valid benefit measure for a country with a high tuberculosis burden. They captured the whole impact of disease on the patients’ health and they allow cross-disease comparisons to be made. Limited information was given on the derivation of the disability weights.
Costs:
The economic analysis was consistent with the perspective stated. The unit costs of the diagnostic tests were reported and were from Indian sources, with plausible ranges of values tested in the sensitivity analyses. These sources were private Indian laboratories and the authors stated that the cost of serology might have been underestimated, favouring this test. The cost of tuberculosis management was from an earlier publication and was presented as a total; this was varied in the analysis of uncertainty. All economic data were treated deterministically. The price year was reported, facilitating reflation exercises for other time periods.
Analysis and results:
The results were clearly presented. The projected costs and benefits were synthesised, using an incremental approach, which allowed the identification of the dominated strategies. Secondary tuberculosis transmission was included, but the model was static and did account for changes, such as repeated rounds of transmission. This was acknowledged as a limitation of the analysis. A deterministic approach was used to assess the uncertainty, and the results were clearly illustrated. The authors stated that their patients had access to laboratories for serological testing, which was not representative of all patients in India. The analysis appears to have been specific to the authors’ context and the results cannot be transferred to other settings.
Concluding remarks:
The methods were valid and transparent, and various areas of uncertainty were considered. The authors’ conclusions seem robust.