Methodological quality of the review: Medium confidence

Author: Vinuesa-Silva JM, Vinuesa-Silva I, Pinazo-Durán MD, Soto-Alvarez J, Delgado-Ortega L, Díaz-Cerezo S

Geographical coverage: Not reported

Sector: Glaucoma

Sub-sector: Conjunctival hyperemia

Equity focus: None specified

Review type: Effectiveness review

Quantitative synthesis method: Meta-analysis

Qualitative synthesis method, if applicable: Not applicable

Background: Glaucoma is one of the most prevalent causes of blindness worldwide. In patients with high intra-ocular pressure (IOP), treatments are based on the reduction of IOP to delay or prevent the progression of the disease. Topical treatments are the most commonly used medications. These drugs may result in local or systematic effects. Conjunctival hyperemia is the most common secondary effect of prostaglandins. Therefore, it is important to systematically assess the literature for this secondary effect.

Objectives: To assess the association of conjunctival hyperemia with the use of a fixed combination of latanoprost/ timolol, through a systematic review and meta-analysis of clinical trials in patients with glaucoma.

Main findings: In this systematic review, eight clinical trials remained for inclusion in the meta-analysis, all of which included patients with ocular hypertension or glaucoma and provided specific hyperemia data for a total of 2600 patients.

There was a moderate heterogeneity (Q: 14.64; df=7; p=0.041; I²=52.2%) and therefore a random effects model was used. The final odds ratio (OR) was 0.47 (CI 95%: 0.24-0.90); p = 0.024. The total conjunctival hyperemia incidence was 2.9% in the latanoprost/timolol group and 7.0% for the competitors (p< 0.0001). The sensitivity analyses indicated that the mean aggregate OR could range between 0.39 and 0.50 depending on the clinical trial excluded of the meta-analysis. None had a significant impact in the global calculation of OR because in all cases the latanoprost/timolol exhibited a degree of hyperemia significantly lower than the comparator group. No bias was found within included studies.

Authors concluded that, compared with other options for the treatment of glaucoma, a fixed combination of latanoprost/timolol is associated with a significant reduction in the development of conjunctival hyperemia. A fixed combination of latanoprost/timolol has the lowest side effects compared with other medications

It was also noted that future studies should include a greater number of trials and search for a longer period of time.

Methodology: The following databases were used in the search for published relevant articles: MEDLINE, EMBASE and Cochrane Controlled Clinical Trials Register between 2000 and 2007. Relevant papers were selected by two independent researchers according to the following criteria:

(a) Comparative clinical trials (excluding crossed designs)

(b) Existence of an active control in the trial

(c) Assessment of the prevalence of conjunctival hyperemia due to treatments

Two authors carried out the extraction of data using a codification sheet. In addition, they evaluated the methodological quality of included studies using the Jadad scale. However, it’s not clear if this was done independently.

Meta-analysis was carried out with the programme Comprehensive Meta-Analysis version 2.2. OR of the various clinical trials were combined using a fixed-effect model of Mantel-Haenszel for homogeneous trials, and the random effect model of Der Simonian and Laird (for heterogeneous clinical trials). Heterogeneity between trials was analysed using Q for Cochrane test and the index /2 calculation which allowed the authors to determine the way in which possible existing heterogeneity could affect the conclusions of the meta-analysis. Publication bias was assessed with the visual inspection of the funnel plot. Sensitivity analysis was conducted to assess the robustness of the results.

Applicability/external validity: The authors did not discuss the applicability/external validity of the results.

Geographic focus: The authors failed to highlight or mention the geographical focus of the review.

Summary of quality assessment: Overall, there is medium confidence in the conclusions about the effects of this study. Authors conducted a search of the literature on relevant databases without any publication type or language restrictions, but restricted the search for studies published between 2000 and 2007.

However, authors did not contact authors/experts and did not check reference lists to ensure that all relevant published and unpublished studies were identified. As such, publication bias should not be excluded from the review. Authors used appropriate methods to select studies for inclusion in the review and to extract data of included studies, minimising bias. Methods used to analyse data of included studies were also appropriate and limitations were acknowledged accordingly.