Selection of comparators (Transition Care Program and usual care) was appropriate. No clear description was provided of the conventional pattern of care.
Data on treatment effect were taken from various sources and included actual data on implementation of the programme, published evidence and authors’ assumptions. Optimistic assumptions for the programme were made and these were not tested in the sensitivity analysis. Clinical data did not appear to be based on solid evidence.
Use of QALYs was appropriate given the potential impact on mortality and morbidity of different length of hospitalisation and readmissions. No details were given on sources of utility weights.
The economic analysis was consistent with the perspective stated by the authors. It appeared that relevant cost categories were included. Economic data were taken from various published sources and some assumptions were required. Unit costs and quantities of resources used were not presented separately as most costs were reported as macro-categories. The cost of the programme was based on a top-down approach and included the entire cost of running a service (including administration, building costs and so on) and might have been overestimated. The Transition Care Program policy stipulated that patients could remain in the programme up to 12 weeks (i.e. more than the expected eight weeks) with a possible inefficiency of the system. The price year was reported, which enabled reflation exercises in other time periods. Variations in cost estimates were not considered.
Analysis and results:
The study results were presented selectively. Cost-effectiveness ratios were not calculated as the analysis focused on cost savings rather than on cost-effectiveness ratios. The cost-utility analysis was based on the hypothetical gains of QALYs associated with the programme. The issue of uncertainty was not investigated. The authors acknowledged potential overestimation in the cost of the programme. Study results were specific to the Australian context and the specific programme analysed.
The study presented some methodological limitations that might affect the validity of the authors’ conclusions.