Interventions:
There were numerous interventions for osteoarthritis of the knee. STOPP allowed continued use of paracetamol and NSAIDs. It was unclear from the study whether STOPP allowed additional treatments such as physiotherapy, exercise and other potential therapies.
Effectiveness/benefits:
The authors acknowledged that one of the primary goals of STOPP was measuring whether chondroitin sulphate structurally modified the knee joint and so potentially prevented later surgery and other treatments. Potential benefits of avoiding surgery or other treatments later were not explored in this study. The utility figures used were well reported from a published mapping methodology with the regression equation reported but they appeared to be an incomplete representation of potential treatment effects over too short a time horizon for a chronic condition such as osteoarthritis.
Costs:
No cost sources were cited so it was impossible to assess the validity of the cost estimates that underpinned the analysis. No costs beyond the cost of the drug were included, costs such as hospital usage, surgeries and other costs should have been evaluated as there may be relevant differences. The authors reported that the STOPP trial showed that chondroitin sulphate may prevent knee surgeries by structurally modifying the cartilage of the knee but this was not represented in the cost-effectiveness analysis presented. The analytical perspective and cost year were not stated so it was difficult to evaluate which potential costs were unrepresented in the study.
The reporting on costs was very poor.
Analysis and results:
The study conducted no sensitivity analysis beyond using minimum and maximum chondroitin sulphate costs from unreported sources and reported no measures of variance on QALYs or costs beyond baseline, so it was impossible to judge what effect uncertainty could have on the analysis. A lack of evaluation of uncertainty made it difficult to evaluate the validity of the results.
The time horizon of the study was only two years yet osteoarthritis of the knee is a chronic lifelong condition. The choice of time horizon for the economic evaluation mirrored the follow-up of the STOPP trial rather than the characteristics of osteoarthritis of the knee. Where a time horizon is not long enough to capture potential effects of treatment on the disease and its progression it is also not possible to judge the validity of the results.
Concluding remarks:
Generally poor reporting, an inadequate time horizon for a chronic condition and no meaningful evaluation of uncertainty made it impossible to determine the validity of the results and the appropriateness of the conclusions.