Interventions:
The interventions were appropriate. The authors gave a description of the intervention and the services provided with it. Specific details of the care delivered before the implementation of RAID were not discussed so the exact nature of the comparator was unclear. The authors did not discuss any other relevant alternatives.
Effectiveness/benefits:
The effectiveness estimates and methods used to derive them were reported clearly. Justification was given for using reductions in readmission and length of stay as surrogates for quality of care and improved health outcomes. Risk of readmission and length of hospital stay were proxy measures of benefit. No direct health measures were evaluated so the real health benefit was unclear. The focus of the paper was evaluating cost savings.
Costs:
The method used to derive the number of bed-days saved was reported clearly. The value of cost applied to one saved bed-day and the method used to derive this were not reported. It was unclear what hospital costs were taken into account when considering the cost of one bed day so it was unclear whether the final reported value of costs saved was appropriate. Costs were assessed from a limited hospital perspective; key hospital costs (such as the cost of the intervention) did not appear to be considered. The effect that early discharge of patients might have had on other NHS resources and on societal costs was not explored.
Analysis and results:
A key limitation of the study concerned the limited scope and reporting of the analysis. It appeared that only certain hospital costs were considered in the analysis. In particular it appeared that the actual cost of the intervention was not included. A full economic analysis should include intervention costs. The results may be valid for the scope of the analysis but the reader should consider carefully whether the analysis captured all costs relevant to their situation.
Another limitation of the analysis was the lack of a full sensitivity analysis; this made it unclear how robust the results were. The alternative LSE results were stated to represent a more conservative analysis but not enough details of this analysis were reported to allow for commentary.
Concluding remarks:
Limitations in the reporting and scope of the analysis, in particular with regards to the costs included, mean that it is not clear whether the authors' conclusions are appropriate.