Methodological quality of the review: High confidence
Authors: Chan EWE, Li X, Tham YC, Liao J, Wong TY, Aung T and Cheng CY
Geographical coverage: None provided
Equity focus: No
Review type: Other review
Quantitative synthesis method: meta-analysis
Qualitative synthesis method: Not applicable
Glaucoma is one of the leading largely-preventable causes of blindness in the world. Surgical treatment aims to open the drainage system and lowering the pressure in the eye when other medical treatments with eyedrops fail. Trabeculectomy is one surgical technique that creates a fistula, allowing drainage of fluid from inside the eye to lower the eye pressure. There are two incision types in this surgery: fornix-based (between the cornea and conjunctiva) and limbal-based (further away under the eyelid).
This review aims to look at whether there are any differences in the surgical outcomes (eye pressure control and complications) between these two different surgical approaches (fornix- and limbal-based techniques).
Six trials met the criteria for inclusion with a total of 361 participants. None of the included trials reported trabeculectomy failure at 24 months. Only one trial reported the failure rate of trabeculectomy as a late complication. Failure was higher among participants randomized to the limbal-based surgery: One in 50 eyes failed trabeculectomy in the fornix group compared with three in 50 in the limbal group (risk ratio (RR) 0.33, 95% confidence interval (95% CI) 0.04 to 3.10); therefore we are very uncertain as to the relative effect of the two procedures on failure rate.
Four studies including 252 participants provided measures of mean IOP at 12 months. In the fornix-based surgeries, mean IOP ranged from 12.5 to 15.5 mmHg and similar results were noted in limbal-based surgeries with mean IOP ranging from 11.7 to 15.1 mmHg without significant difference. Mean difference was 0.44 mmHg (95% CI −0.45 to 1.33) and 0.86 mmHg, (95% CI −0.52 to 2.24) at 12 and 24 months of follow-up, respectively. Neither of these pooled analyses showed a statistically significant difference in IOP between groups (moderate quality of evidence).
One trial reported number of anti-glaucoma medications at 24 months of follow-up with no difference noted between surgical groups. However, three trials reported the mean number of anti-glaucoma medications at 12 months of follow-up without significant difference in the mean number of postoperative IOP-lowering medications between the two surgical techniques. Mean difference was 0.02, (95% CI −0.15 to 0.19) at 12 months of follow-up (high quality of evidence).
Because of the small numbers of events and total participants, the risk of many reported adverse events was uncertain and those that were found to be statistically significant may have been due to chance.
For risk of bias assessment: although all six trials were randomized selection bias was mostly unclear, with unclear random sequence generation in four of the six studies and unclear allocation concealment in five of the six studies. Attrition bias was encountered in only one trial which also suffered from reporting bias. All other trials had an unclear risk of reporting bias as there was no access to study protocols. All included trials were judged to have high risk of detection bias due to lack of masking of the outcomes. Trabeculectomy is quite a standard procedure and unlikely to induce bias due to surgeon ‘performance’, hence performance bias was not evaluated.
The authors included all randomized controlled trials (RCTs) assessing the beneficial and harmful effects of fornix- versus limbal-based trabeculectomy for glaucoma, irrespective of publication status and language. Participant aged 18+ with IOP 21mmhg were included. Any type of glaucoma was considered, participants who had undergone trabeculectomy already with at least 12 months of follow-up were included. Interventions included fornix-based trabeculectomy versus limbal-based trabeculectomy with or without adjunctive use of MMC and concomitant performance of cataract extraction.
The authors was conducted on CENTRAL (which contains the Cochrane Eyes and Vision Trials Register) (2015, Issue 9), Ovid MEDLINE, Ovid MEDLINE In-Process and Other Non-Indexed Citations, Ovid MEDLINE Daily, Ovid OLDMEDLINE (January 1946 to October 2015), EMBASE (January 1980 to October 2015), Latin American and Caribbean Health Sciences Literature Database (LILACS) (January 1982 to October 2015), the ISRCTN registry (www.isrctn.com/editAdvancedSearch), ClinicalTrials.gov (www.clinicaltrials.gov) and the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) (www.who.int/ictrp/search/en). No date or language restrictions in the electronic searches for trials.
Two authors independently extracted data using a form developed by Cochrane Eyes and Vision. If they could not reach agreement, they listed the study in the ’Studies awaiting classification’ table until further clarification was available from the authors. In case of more than one publication of the same RCT, we reviewed data from all articles and extracted data as appropriate. The two authors resolved disagreements by discussion. Two authors independently assessed the included studies for sources of bias according to the guidelines in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). Study investigators were contacted in case of missing data. Clinical and methodological heterogeneity of included studies was assessed by examining variations in the study design and methods, characteristics of the participants, variation in interventions, and length of follow-up. It was also assessed by statistical heterogeneity in each meta-analysis using the I2 and Chi2 statistics. We regarded statistical heterogeneity as substantial when either the I2 statistic was greater than 50% or there was a low P value (< 0.10) in the Chi2 test for heterogeneity. Sensitive analysis was planned to determine the impact of excluding studies with high risk of bias.
The authors mentioned that with only six trials in this review, the evidence of the difference in efficacy and complications between fornix versus limbal trabeculectomy was incomplete. It was uncertain in this review whether there was a difference between the two surgical techniques in lowering IOP at 12 and 24 months of follow-up.
With just six studies and no studies from LMIC, is difficult to conclude that this review can be apply in on low/ middle income countries.
Summary of quality assessment:
High confidence was attributed in the conclusion of this review. The authors used rigorous methods to conduct the review, ensuring biases were avoided at all stages of the review.
Al‐Haddad C, Abdulaal M, Al‐Moujahed A and Ervin AM (2015) Fornix‐based versus limbal‐based conjunctival trabeculectomy flaps for glaucoma. Cochrane Database of Systematic Reviews, (11).