Methodological quality of the review: Low confidence
Author: Orme M, Boler A
Geographical coverage: United States of America (USA) and Japan.
Sub-sector: Prostaglandin analogues, primary open angle glaucoma, treatment, cost-effectiveness
Equity focus: Patients with open angle glaucoma or ocular hypertension
Review type: Effectiveness review
Quantitative synthesis method: Narrative synthesis
Qualitative synthesis method: Not applicable
Glaucoma is the second leading cause of blindness globally. Preventing progression to blindness preserves a patient’s quality of life (QoL), but requires long-term resource-intensive treatment and the full commitment of the patient. The European Glaucoma Society (EGS) noted that careful evaluation of the cost of glaucoma management was needed, including cost to the individual and to society. Furthermore, EGS updated its recommendation to reflect a shift from beta blockers to prostaglandin analogues (PAs) as first-choice therapy. It is in this context that this systematic review is placed.
The aim of this systematic review was to identify and critically appraise the cost-effectiveness evidence for PAs in the treatment of glaucoma and, more specifically, primary open angle glaucoma (POAG), and how well this evidence informs decision-makers.
Main findings: This section should include a brief description of findings, including number of included studies, location of studies, results of synthesis (if meta-analysis was conducted, include pooled effect size and 95% confidence interval for relevant syntheses), conclusions, findings regarding methodology and future research.
The authors included 13 cost-effectiveness full-text studies and 13 abstracts in the systematic review. The reviewed studies covered the US and European settings, with one Japanese study included. The literature mainly focused on identifying the direct cost of glaucoma treatment in the short term, and particularly the precise quantification of glaucoma drug costs.
However, as authors note, studies included in the review were very diverse because of inconsistency in the outcome measures. Some studies were not regarded as true economic evaluations as these were cost-identification studies, and other studies did not report comparative results. Additionally, few studies may have included patients without confirmed glaucomatous damage. Therefore, authors note that little overlap was seen between independent studies in terms of patient group, line of treatment or disease stage. As such, authors conclude that ‘as a body of evidence, it is insufficient in providing the range of cost-effectiveness evidence needed to support treatment choices involving PAs’.
Overall, studies included were methodologically flawed. Major issues were in the measurement of clinical effectiveness, and no single study had a measure that was both clinically relevant and calculated in a robust and way. Other studies lacked transparency and methodological rigour, and as such results were unreliable.
Some studies did not specify an intra-ocular pressure (IOP) target in the definition of control and as such they do not differentiate between aggressive or conservative treatment targets. In studies that did allude to a specific target, it was not clear whether the target was selected to fit the study population.
Authors included economic evaluations comparing different types of prostaglandin analogues interventions (e.g. latanoprost) in patients with open angle glaucoma or ocular hypertension. Authors searched MEDLINE, EMBASE and the UK NHS economic evaluations databases up to July 2005, but limited searches to studies written in English only.
To supplement the studies retrieved from the literature databases, abstracts from the following conference proceedings were hand searched: International Society for Pharmacoeconomics and Outcomes Research, Association for Research in Vision and Ophthalmology, Academic of Managed Care Pharmacy, and the American Academy of Ophthalmology. Abstracts of articles retrieved by the searches were inspected and included according to pre-defined inclusion criteria. Two reviewers assessed each paper and extracted data of study and a third reviewer reconciled any differences in opinion.
The studies were also critically appraised using a checklist and grading system developed by the authors for this review. The checklist was based on the Scottish Intercollegiate Guidelines Network (SIGN) Methodology checklist and consisted of series of questions that aimed to establish the internal validity of the study as well as relevance of the study and economic implications.
The authors used a narrative synthesis approach to review the literature, which seems appropriate due to the diversity of included studies.
Authors note that the scope of the studies retrieved from the literature search was wide ranging. The lack of comparability meant that authors were unable to assess the convergent validity of the results or make any recommendations regarding the relative cost and consequences of PA use.
The review authors did not report restricting the literature search to specific income settings. As studies included were not reliable, authors did not draw conclusions or recommendations regarding the cost of using PA.
Summary of quality assessment:
Low confidence was attributed in conclusions about the effects of this study as important limitations were identified in this systematic review. Literature search was not comprehensive enough that we can be confident that relevant studies were not omitted in the review, as authors did not avoid language bias and did not contact authors/experts for further relevant studies. Potentially, this indicated the presence of publication bias within the review.
Authors used appropriate methods to analyse the findings and clearly described heterogeneity among included studies, however methods used to extract data were clearly not reported. Additionally, it is not clear if the two reviewers critically appraised included studies independently. Although authors do not draw strong conclusions based on studies included in the review, these limitations may impact on the overall validity of the review.