Interventions:
The selection of the comparators was appropriate as the two available preventive strategies were considered.
Effectiveness/benefits:
An appropriate approach was used to identify the relevant sources of evidence. The published literature was searched and the details of this review were clearly presented. The quality of studies and the validity of their findings were assessed using the Jadad scale for randomised controlled trials, and the Oxman-Guyatt index for the meta-analysis. This meta-analysis attained a maximum score, suggesting a very high quality of evidence; only head-to-head randomised controlled trials were analysed. The key clinical parameters were varied in the sensitivity analysis. The rate of heterotopic ossification was an intermediate outcome of the interventions. The authors pointed out that it did not capture the potential complications of treatment, but these were found to be similar between the two treatments. The authors acknowledged that an assessment of quality of life would have been more appropriate, but utility estimates were not found in the literature.
Costs:
The perspective was not explicitly stated, but the costs were relevant to the health care setting and limited to the period after surgery. These costs were from the USA, but costs from European countries were also considered and showed similar differences between the two interventions. The key unit costs were provided, allowing the replication of the analysis for other settings. The price year was not explicitly stated, but the costs referred to 2007 data. The impact of varying the cost estimates was extensively investigated in the sensitivity analyses.
Analysis and results:
The results were clearly presented. An incremental approach was used to synthesise the costs and benefits of the two strategies. The details of the decision model were given. The uncertainty was satisfactorily investigated, considering the variability in the key inputs. The authors stated that although the costs were mainly from the USA, the large difference between radiation and NSAIDs was likely to exist in other contexts, making the findings transferable. It was unclear if a longer time horizon might have changed the cost-effectiveness results.
Concluding remarks:
The methods were valid and key areas of uncertainty were considered. The authors’ conclusions appear to be robust.