Methodological quality of the review: Low confidence
Author: Lee CW, Buckley F, Costello S, Kelly S
Geographical coverage: United States of America (USA) and Europe
Sub-sector: Glaucoma treatment
Equity focus: None specified
Review type: Other review
Quantitative synthesis method: Narrative/thematic synthesis
Qualitative synthesis method: Not applicable
Background: Glaucoma is one of the leading causes of blindness caused by high intra-ocular pressure (IOP), for which the management is to reduce IOP generally by laser surgery, the administration of eye drops or systemic medication. Of the pharmacological treatments available (beta blockers, prostaglandins, carbonic anhydrase inhibitors and miotics), prostaglandins form the mainstay of therapy.
A large number of randomised controlled trials (RCTs) have been published examining the efficacy of pharmacological treatments on glaucoma. An evidence-based-medicine approach to reviewing the characteristics of the existing RCTs will be helpful to enable a focus on clinically meaningful issues, and to provide balanced views to aid clinical and public health decision-making.
Objectives: To classify the comparative quantity and quality of the RCT evidence of each pharmacological treatment for glaucoma.
Main findings: In this review, 510 publications were identified, of which 181 studies had a prostaglandin treatment arm. The median study duration was 12 weeks (IQR 4-13) and 78% of included trials had a duration of three months or less. Only four studies had a duration over one year and included a latanoprost and timolol treatment arm.
The geographical location of studies was generally distributed evenly across treatments and the majority of studies were conducted in the US and Europe. From the reported data, latanoprost is the only prostaglandin with UK-specific studies (12-14) and Asia-specific studies (N=13). Authors noted that, overall, included trials were poorly reported. In the majority of cases, information that would contribute to the Jadad score was not reported.
Based on studies included in the review, authors noted that there was a lack of data on younger populations (median of the mean ages of included patients was 63.4 years [IQR: 61–66 years]). Caucasians constituted 79.6% of the studied population. Evidence by ethnicity, as well as by co-morbidity, was scarce. For 92% of studies, the primary outcome was IOP reduction; little was reported on other indicators. The most commonly reported adverse effect was hyperemia. Authors noted that well-conducted observational studies have to be considered in conjunction with RCTs to provide a more complete and deeper understanding of this topic.
Methodology: The following databases were searched for relevant studies: MEDLINE, EMBASE and Cochrane CENTRAL, and conference proceedings were conducted up to March 2007. Other relevant databases were also searched. All RCTs recruiting adults with POAG and/or ocular hypertension (OH) were included. The treatments considered were UK licensed POAG/OH drugs given as mono-therapy or in combination. Articles not in the English language were excluded.
Screening of studies for inclusion were conducted by two reviewers independently. Extraction of studies used a standardised form within the Access database, and was conducted by two reviewers independently. A critical appraisal (using the Jadad score) was also undertaken by two reviewers independently.
The characteristics of the included studies were analysed and summarised narratively. The results from this systematic review were subsequently used to determine key comparisons for further detailed extraction to be undertaken.
Applicability/external validity: Authors did not discuss the applicability/external validity of findings. However, authors noted that included studies were mainly from the US and Europe, and from the reported data latanoprost is the only prostaglandin with UK-specific and Asia-specific studies.
Geographic focus: The majority of studies were conducted in the US and Europe.
Summary of quality assessment: Authors used a narrative synthesis which seemed appropriate due to the diversity of the included studies. Authors used appropriate methods to screen studies for inclusion and to extract data of included studies, and used appropriate tools to assess the risk of bias of included studies.
However, major limitations were identified in the review. Authors did not conduct a thorough search of the literature as references of included studies were not reviewed and authors/experts were not contacted as part of the search strategy. Language bias was not avoided as studies written in English only were included in the review. This may imply the presence of publication bias within the review.
Authors did not address the extent of heterogeneity within the review. In terms of data synthesis, the outcome measure of IOP was not consistent across the studies and authors acknowledged the lack of homogeneity in the reported outcomes. However, the likelihood of bias within the included studies was not clearly addressed. Therefore, low confidence was attributed in the conclusions about the effects of this study.