Interventions:
The rationale for selection of the comparators was clear. Pre- and post-intervention periods were compared to assess the efficacy of the screening programme. Limited information was given on the conventional pattern of care prior to the active surveillance programme.
Effectiveness/benefits:
The clinical analysis was subject to potential limitations such as the retrospective nature of the study and limited information on the clinical and demographic characteristics of the pre- and post-intervention groups of patients. Another critical issue was the lack of contemporaneous assessment of outcomes: as the two patient samples were considered over two different periods of time, the impact of time-related bias could not be ruled out because factors other than the study intervention might have affected the endpoints. No statistical analysis was conducted to adjust clinical results for potential differences in patient groups. Evidence came from a single medical institution which might not be representative of other health care systems. The endpoints used in the analysis represented the natural outcomes of the surveillance programme but were intermediate measures of the impact of the interventions on patients’ health.
Costs:
The economic analysis was performed from the economic viewpoint of the hospital. Details of unit costs and resource quantities for the surveillance programme were reported extensively and this enhanced the transparency of the economic side of the study. The value used for MRSA infection was the lowest found in the literature in order to be conservative against the surveillance intervention. Methodological details were not provided for the selected study. The price year was not stated so reflation exercises in other time periods was not possible. The impact of variations in economic estimates was not investigated.
Analysis and results:
The study results were presented extensively. Cost-effectiveness ratios were not calculated because of the cost-consequences framework of the analysis. No sensitivity analyses were carried out to deal with the issue of uncertainty and this represented a limitation of the analysis. The authors highlighted the difference found between surgical and medical intensive care units and explained possible reasons. The authors acknowledged some potential drawbacks of the analysis (such as use of census-days instead of days at risk) but did not underline the issues related to the use of a clinical study with potential bias.
Concluding remarks:
The analysis was well-presented but had some methodological limitations. The authors’ conclusions should be treated with caution.