Interventions:
The interventions were adequately described and appeared to be appropriate comparators. It was likely that all the relevant comparators were included in the study (25 combinations were tested); a no-intervention strategy was appropriately included. Given the number of comparators included, it was likely that the interventions would be generalisable to other study settings.
Effectiveness/benefits:
A significant proportion of the effectiveness data were taken from the effectiveness data identified in the previously published simulation model used. However, it was unclear if systematic methods were used to identify the effectiveness data in this previous study. It was unclear if systematic methods were used to derive the remaining effectiveness data in this study. Therefore, it was unclear if the best available evidence was included in the study. The authors stated that the data on the natural history of the disease were based on the best available data regardless of the setting, but the model was calibrated to the Thailand context using country-specific epidemiological data. However, the evidence included and the techniques used were not described and so their quality was unclear.
The benefit measure of life years was valid, as it captured the impact of the disease on the most relevant dimension of health, which was survival. However, life years did not take account of morbidity, whereas measures such as quality-adjusted life-years would do this and may be of interest for the study interventions.
Costs:
The study perspective was clearly stated; the cost categories included appeared to appropriately cover this perspective. Limited information on the methods of the cost analysis and sources of cost and resource data were provided, so their quality and appropriateness was unclear. Additional information on the cost analysis was provided in a separate online appendix. The costs appear to have been appropriately discounted and adjusted for inflation.
Analysis and results:
The analytical approach was adequately described; additional details were provided in the online appendix. Adequate sensitivity analysis was undertaken to assess the uncertainty of the study results. However, the overall model uncertainty could have been better assessed with probabilistic sensitivity analysis. The life years saved and costs of each of the interventions were not reported. Although selected incremental results were reported, this seemed appropriate given the large number of strategies being compared. The authors discussed several limitations with their study.
Concluding remarks:
The analytical approach of the study appears appropriate. Although some of the methods and results were poorly reported, the authors conclusions appear to be appropriate and reflect the evidence available.