The authors did not specify how the sample or sample size was selected and how bias might have been avoided. No justification was provided for the type of study design used. A retrospective case series may not be the most appropriate design, as the sample may have been biased. However, given the clear finding that CT examination was not very effective in detecting serious cases, and was very expensive, this may not alter results drastically. Cost per case of important pathology is an appropriate measure to use in the economic analysis of this intervention. However, there was no proper economic analysis of the comparator, (do nothing). It was assumed that serious cases would present themselves with symptoms which would mean they would be detected anyway. The costs of failing to find a serious case under this strategy were not considered. The hospital perspective used was probably the most useful. Additional costs to patients due to visits for CT examination should have perhaps been considered: it is assumed the authors considered these to be insignificant. As the authors stated, analysis was likely to be sensitive to the cost of CT examination. As no sensitivity analysis was performed generalisabilty of results to other settings/countries may be difficult given price variation. Opportunity cost was mentioned indirectly in emphasising the budget constraint. The methodology of the economic analysis was clearly not rigorous, however it is probable that similar results would be achieved with a more thorough approach. However, only a proper economic analysis comparing alternatives will reveal this.