The intervention appeared to be appropriate. The intervention and comparator were clearly and comprehensively reported and the most appropriate comparator (standard care) was used in the analysis.
Details of how and when patients' VAS scores were derived were reported clearly. A diagram of the VAS used was supplied. Details of how and when surgeons were asked to rate the ease of the procedures was not reported. Details of operative complications were not reported clearly; specific numbers were reported only for subconjunctival haemorrhage. Effectiveness results for the primary outcomes were reported clearly. All of the outcome measures could be regarded as subjective but are commonly used to assess perceived change in patients. Some of the results achieved were statistically significant but it was not clear that the study was powered to detect significant differences.
Costs were reported clearly. It appeared that only the cost of the intervention was included in the cost estimate; this could be considered limiting. The authors stated that less time was required to perform the procedure using the conjunctival mould so more injections would be able to be performed in a session, providing further financial benefits; this was not fully assessed or supported by evidence. However, assuming that the two procedures did not differ in terms of other resources needed (such as additional training or future complications) the authors' conclusion that the conjunctival mould would lead to cost savings appeared reasonable.
Analysis and results:
The authors highlighted the need for a larger study to fully establish the clinical utility of the intervention. Baseline characteristics for each of the patient cohorts were not reported so it was not possible to assess whether the patient populations were comparable at baseline and whether key prognostic factors differed. No attempt was made to control for any potential confounding factors in their analysis so the risk of bias was likely to be significant. There was no sensitivity analysis to assess the impact of parameter uncertainties so it was not possible to draw any conclusions regarding the robustness of the results.
Overall the reporting was satisfactory but methodologically the risk of bias was high. The recommendation for a larger study was warranted. The cost-effectiveness results presented should be interpreted cautiously.