Selection of comparator:
The reason for the choice of comparator (acute rehabilitation) is clear.
Internal validity of estimate of benefit:
The internal validity of the estimates of benefits cannot be guaranteed due to the retrospective nature of the study design coupled with the possibility of non-comparability of patients and selection bias, (as acknowledged by the authors).
Internal validity of estimate of costs:
Quantities of length of stay and average daily billed treatment hours were reported separately from the costs. Insufficient details of the methods of cost estimation were given. The use of charge data as opposed to true costs, coupled with the retrospective nature of the cost analysis, may, as the authors acknowledge, have adversely affected the study's internal and external validity.
Other issues:
In view of the retrospective nature of the study design and the lack of sensitivity analysis, the study results need to be treated with some degree of caution. The authors acknowledged that, since it is not clear how representative the study populations and treatment methods were of a larger sample of facilities, given the rapidly changing treatment scene, the applicability and generalisability of the study results to other settings cannot be guaranteed. Appropriate comparisons with other studies were not possible since, at the time of the study, the concept of subacute rehabilitation was new and it was difficult to find any scientific literature on this subject.