The authors did not provide an explicit justification for their selection of the comparators. Travoprost was the most commonly prescribed first-line therapy over the study period (2002 to 2005) and the comparator contained two of the most commonly used therapies for this patient population.
The clinical analysis was based on data from an administrative database. The main advantages of this database were that it represented the general population well, due to the large number of GPs and patients involved, and it represented real-world treatment patterns. The main drawback was the fact that data were retrospectively analysed. The authors stated that prescription renewals were not clearly documented and clinical information had to be extrapolated to a form suitable for the clinical analysis. A statistical analysis was necessary to balance the two samples of patients for time to diagnosis, which was significantly shorter in the travoprost group. Other differences in the frequency of ocular co-morbidities were also shown. Patients in the travoprost group generally appeared to have a lower health status, but none of these conditions were directly associated with the time to treatment failure. These issues should be considered when judging the internal validity of this analysis.
The analysis of costs was consistent with the economic viewpoint of the UK payer. The cost items were presented, but surgery and laser strategies were reported as macro-categories, without unit costs. The resource quantities were presented in detail. The use of statistical tests enhances the validity of the economic analysis. The sources of data were reported for most items, but some costs were presumably calculated on the basis of authors’ opinions, as their sources were not reported. In general the economic analysis was conducted satisfactorily.
Analysis and results:
The costs and benefits were not synthesised and, in effect, a cost-consequences analysis was conducted. The issue of uncertainty was not investigated, which was a limitation of the study. The authors acknowledged some limitations of their analysis, mainly related to the design of the clinical study. The results of this study were also compared with a published randomised controlled trial, which had shown different findings.
The study had some methodological limitations that may affect the validity of the authors’ conclusions.