Description of the condition
A systematic review of all population-based surveys on blindness and low vision by the World Health Organization (WHO) in 2002 estimated that 37 million people are blind worldwide, with 12.3% (4.4 million) attributable to glaucoma, second only to cataract (48%) (Bourne 2006). Quigley et al project that 8.4 million people will be blind from primary glaucoma by 2010, rising to 11.1 million by 2020. The numbers who are blind are a fraction of those with the disease; the authors estimate the combined number of people with primary glaucoma to be 60.5 million by 2010, increasing by 20 million over the subsequent decade (Quigley 2006). Nevertheless, it has also been estimated that half of all people with glaucoma do not know that they have it and are, therefore, not receiving treatment that may prevent vision loss. Studies consistently report a prevalence rate for primary open-angle glaucoma (POAG) of 1% to 2% of white adults. However, significant racial differences exist with higher rates in dark races (Tielsch 1991). In addition, the Baltimore Eye Survey described that the rates of visual impairment among African-Americans were twice that of whites (Tielsch 1991).
Presentation and diagnosis
Primary open-angle glaucoma, even under treatment, has been associated with measurable rates of progression of visual field loss. Published studies indicate that on an annual basis between 3% and 8% of patients under the care of an ophthalmologist may suffer progressive field damage (Kass 1976; Mao 1991; Quigley 1996). A diagnosis of open-angle glaucoma is made on the basis of a combination of clinical signs including: intraocular pressure (IOP) measurement, a clinically open angle identified by gonioscopy, optic disc (asymmetry of cup/disc ratio more than 0.2 or glaucomatous pathological disc, ranging from notching to advanced cupping, or both) and classical glaucomatous visual field changes.
Description of the intervention
Options for management include medical treatment, laser therapy and surgical intervention. In general, filtration surgery is indicated when medical and laser therapies are insufficient to control the glaucoma, and when the rate of deterioration of visual function is rapid enough to damage the patient's quality of life (Spaeth 2000). Patients with open-angle glaucoma are operated on in order to increase the outflow of aqueous. Trabeculectomy is a filtering surgery in which removal of a full-thickness block of eye filtration tissue is done to achieve decreased resistance to the outflow filtration of aqueous (eye fluid that contributes to eye pressure) and subsequently lowering of eye pressure. It is considered by many ophthalmologists to be the gold-standard glaucoma operation. However, it is associated with significant postoperative complications such as hyphaema, shallow or flat anterior chamber, hypotony, choroidal detachment and hypotony maculopathy, all due to excessive filtration, and subsequent development of cataract. A new approach in trabecular surgery has been developed to minimise these complications; this is non-penetrating filtering (trabecular) surgery (partial-thickness removal of tissue) (Mortensen 2004). There are two widely practised technical approaches to non-penetrating filtration surgery (NPFS). Deep sclerectomy involves the creation of conjunctival and scleral flaps similar to a trabeculectomy; a deeper inner block of scleral tissue is excised under the scleral flap creating a Descemet's window that allows aqueous seepage from the anterior chamber. Subsequent fluid percolation proceed subconjunctivally, resulting in a filtration bleb. Further placement of a collagen implant in the scleral bed has been reported to maintain the sub-scleral space (Sanchez 1997;Tan 2001). In the second technique, viscocanalostomy, a high viscoelastic material is injected through the two open ends of Schlemm's canal to dilate it. The superficial scleral flap is sutured so tight and viscoelastic is injected beneath the scleral flap at the end of the intervention prior to closure of conjunctiva (Guedes 2006; Hamard 2002; Stegmann 1999).
How the intervention might work
Surgery is an effective way to lower IOP (Burr 2012). Bylsma hypothesises that if the safety margin of glaucoma surgery could be increased significantly without sacrificing efficacy, surgical intervention for glaucoma might be considered earlier (Bylsma 1999). Zimmerman et al reported favourable results of non-penetrating trabeculectomy in phakic and aphakic patients (Zimmerman 1984). Stegmann et al described a similar technique in which the scleral space is filled with a viscoelastic substance. They reported a complete success rate of 82.7% and a qualified success rate of 89.0% over a 35-month follow-up (Stegmann 1999). Fyodorov as well as Kozlov et al, described placing a collagen implant in the scleral bed to enhance the filtration of deep sclerectomy (Fyodorov 1990; Kozlov 1990). Sanchez and co-authors also reported a better surgical outcome when the collagen implant is used (Sanchez 1997). Chiou et al reported ultrasonic biomicroscopy findings consistent with IOP-lowering by aqueous filtration through the thin remaining trabeculo-descemetic membrane (TDM) to an area under the scleral flap, which was hypothetically kept open by the presence of the collagen implant (Chiou 1998). Other available implants are the reticulated hyaluronic acid implant, SKGEL implant and the hydrophilic acrylic non-absorbable implant.
Why it is important to do this review
Although conventional trabeculectomy has been considered the optimum approach for IOP reduction, the high possibility of both early and late related complications directs the interest in evaluation of non-penetration glaucoma surgery as a developing new surgical procedure. A systematic review is needed to evaluate the effectiveness of the new procedure and its potential for fewer complications and greater acceptability to patients.
The objective of this review is to examine the effects of non-penetrating trabecular surgery (viscocanalostomy or deep sclerectomy with or without adjuvants) compared with conventional trabeculectomy (modified Cairns-type technique), when used to treat people with open-angle glaucoma.
Criteria for considering studies for this review
Types of studies
We included relevant randomised controlled trials (RCTs) and quasi-RCTs. A quasi-randomised trial is one that uses quasi-randomisation to allocate participants to different interventions. Quasi-randomisation is a method of allocating participants to different forms of care that is not truly random; for example, allocation by date of birth, day of the week, medical record number, month of the year or the order in which participants are included in the study. We included studies which gave outcome data at a minimum of 12 months.
Types of participants
Participants in the trials were people with open-angle glaucoma who had undergone the surgical treatments in question. There was no restrictions on age, gender or ethnicity.
Types of interventions
We included trials comparing non-penetrating filtration surgeries (NPFS) (viscocanalostomy and deep sclerectomy) with conventional trabeculectomy. Antimetabolites may have been used in either or both arms of the trials.
Types of outcome measures
Proportion of successful procedures at least 12 months after surgery. Intraocular pressure was as measured by applanation tonometry in each included trial. Total success was defined as a target intraocular pressure (IOP) at 12 months or more post surgery being less than 21 mmHg without additional topical IOP-lowering medications. Partial success was defined as pressure at 21 mmHg or below with or without medication.
- Progressive visual field loss according to the criteria defined in the methodology of each trial. We described the instrument used to quantify visual field loss and the definitions of progressive visual field loss for each included study, whenever possible.
- Progression of optic disc damage or nerve fibre layer loss according to the criteria defined in the methodology of the trial.
- Reduction of LogMAR score equal to or greater than 0.3 approximating to a Snellen visual acuity of 2 lines or more.
- Quality of life measures, including whether or not there had been a reduction in use of IOP-lowering medications following surgical interventions.
The secondary outcome measures were measured at one year.
Any adverse effects related to the interventions. Complications following surgery include: hypotony, wound leak, infection, cataract progression and cataract surgery. We also recorded the number of cases where NPFS had to be converted to trabeculectomy.
Search methods for identification of studies
We searched the Cochrane Central Register of Controlled Trials (CENTRAL) (which contains the Cochrane Eyes and Vision Group Trials Register) 2013, Issue 8, part of The Cochrane Library. www.thecochranelibrary.com (accessed 27 September 2013), Ovid MEDLINE, Ovid MEDLINE In-Process and Other Non-Indexed Citations, Ovid MEDLINE Daily, Ovid OLDMEDLINE (January 1946 to September 2013), EMBASE (January 1980 to September 2013), Latin American and Caribbean Health Sciences Literature Database (LILACS) (January 1982 to September 2013), the metaRegister of Controlled Trials (mRCT) (www.controlled-trials.com), ClinicalTrials.gov (www.clinicaltrials.gov) and the WHO International Clinical Trials Registry Platform (ICTRP) (www.who.int/ictrp/search/en). We did not use any date or language restrictions in the electronic searches for trials. We last searched the electronic databases on 27 September 2013.
See: Appendices for details of search strategies for CENTRAL (Appendix 1), MEDLINE (Appendix 2), EMBASE (Appendix 3), LILACS (Appendix 4), mRCT (Appendix 5), ClinicalTrials.gov (Appendix 6) and the ICTRP (Appendix 7).
Searching other resources
We searched the abstracts of the Association for Research in Vision and Ophthalmology (ARVO) for the years 1988 to 2007 using keywords:
- non-penetrating glaucoma;
- deep sclerectomy/viscocanalostomy/trabeculectomy;
- collagen implant/SKGEL/reticulated hyaluronate implant/polymegma/mitomycin C/5-fluorouracil;
Data collection and analysis
Selection of studies
Two authors independently reviewed the titles and abstracts of all studies identified electronically and by handsearching. We obtained full copies of all potentially eligible studies and assessed each according to the definitions in the 'Criteria for considering studies for this review' section. We resolved disagreements by discussion between the review authors. Where necessary we attempted to obtain additional information from the principal investigators of the trials. We arranged for translation of trials published in a language other than English.
Data extraction and management
Two review authors independently extracted data onto a modified version of a form developed by the Cochrane Eyes and Vision Group. When data were missing or difficult to determine from the paper, we contacted the authors for more information. The review authors compared the extracted data and resolved discrepancies by discussion. We extracted the following information.
- Methods: methods of allocation, masking (outcome assessment), exclusions after randomisation, losses to follow-up, compliance and study design, intention-to-treat or available case analysis.
- Participants: country of enrolment, number randomised, age, sex, ethnicity, main inclusion and exclusion criteria.
- Interventions: type of surgical method, use of adjuvants, any immediate (within two weeks) postoperative interventions.
- Outcomes: we collected data on all identified outcomes together with length of follow-up and exclusions/drop outs.
- Data entry: two review authors collected the data in spreadsheets then entered the data into Review Manager 5 (RevMan 2012).
Assessment of risk of bias in included studies
Two review authors independently assessed each included study for risk of bias according to chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We considered five parameters of quality.
- Adequate sequence generation: Was the sequence of allocation of participants to groups adequately generated?
- Allocation concealment: Was the sequence of allocation of participants to groups concealed until after the interventions were allocated?
- Blinding (masking): Were the persons assessing outcome unaware of the assigned intervention?
- Incomplete outcome data: Were rates of follow-up and compliance similar in the groups? Was the analysis by intention-to-treat, i.e. were all participants analysed as randomised and were all randomised particpants analysed?
- Selective reporting of outcomes: Are the reports of the study free of suggestions of selective outcome reporting?
We assessed each question-based entry as 'low risk' of bias, 'high risk' of bias or 'unclear' and this is presented in the 'Risk of bias' table for each study included. We contacted study authors for clarification if any parameter was considered 'unclear'. We included trials considered as 'high risk' of bias on any parameter in the analysis, however, we assessed the effect of excluding these trials in a sensitivity analysis. We did not grade trials according to performance bias (masking of participants and researchers) as the trial participants and persons providing care could not be masked. However, bias could be reduced by using observers masked to the intervention when assessing the primary outcome.
Measures of treatment effect
Our primary outcome was dichotomous and thus our measure of treatment effect was the odds ratio. We compared the odds of total success (target pressure without drops) between treatment groups and the odds of control with or without drops between treatment groups.
Unit of analysis issues
We included studies which had used one eye per participant and those which used two eyes per participant, but we took account of pairing using the generic inverse variance method in our analysis. Paired studies were entered as clustered trials and an effect estimate computed. Studies with a single eye were entered and an effect estimate computed. These effect estimates were then meta-analysed using the generic inverse method.
Dealing with missing data
We attempted to extract data from the papers to enable an available case analysis. We noted the proportion of participants who did not provide outcome data in the study characteristics table. If drop outs were very high or were different across treatment groups then we assessed that study as at 'high risk' of bias and excluded it from the meta-analysis but not from the review. In the case of missing data, we used an available case analysis method.
Assessment of heterogeneity
We carefully reviewed the trial reports to identify clinical diversity. We used the Chi
Assessment of reporting biases
We planned to use funnel plots to assess for publication bias, however, since there were fewer than 10 studies identified by our review this was not possible. Should future trials become available we will use funnel plots to assess reporting bias.
We used the random-effects model since we believe that our studies estimate effects which follow a distribution across studies. If there were fewer than three trials (i.e. limited data) we did consider use of a fixed-effect model. If high (I
- deep sclerectomy versus conventional trabeculectomy;
- viscocanalostomy versus conventional trabeculectomy.
Subgroup analysis and investigation of heterogeneity
We conducted subgroup analysis for the type of NPFS: sclerectomy or viscocanalostomy.
We did not conduct a sensitivity analysis to assess the effect of excluding trials assessed as 'high risk' on any aspect of trial quality due to a small number of trials being identified but will do so in future updates of this review.
Description of studies
Results of the search
The electronic searches yielded a total of 1756 references (Figure 1). After deduplication the Trials Search Co-ordinator scanned 1158 records and removed 1044 references which were not relevant to the scope of the review. We screened the remaining 114 references against the 'Criteria for considering studies for this review'. We obtained full-text reports of 23 citations for further investigation. We contacted six trial authors via their corresponding emails: Chiselita 2001; Cillino 2005; El Sayyad 2000; Kobayashi 2003; Russo 2008 and Yalvac 2004. For Kobayashi 2003 and Yalvac 2004 we received invalid email reply messages. With the exception of Russo 2008, no other trial authors replied to our emails. We found five studies meeting the inclusion criteria. A summary of the characteristics of the included studies is given below. The other 18 studies were excluded for various reasons.
|Figure 1. Results from searching for studies for inclusion in the review|
We included five trials in the review (Cillino 2005; El Sayyad 2000; Kobayashi 2003; Russo 2008; Yalvac 2004). Evidence of quasi-randomisation was found in Cillino 2005 and Russo 2008. Full details of the trials can be found in the 'Characteristics of included studies' table.
Setting, participants and interventions
Three of the five trials were conducted in Europe, two in Italy and one in Turkey, and two were carried out in Asia, Japan and Saudi Arabia. Three hundred and five eyes with POAG were included across all five studies. In addition, Cillino 2005 had included six eyes with pseudoexfoliative glaucoma as part of their work. A total of 311 eyes from 247 participants were included in this review in which 133 eyes (participants) were quasi-randomised. One hundred and fifty-one eyes had non-penetrating glaucoma surgery (101 deep sclerectomy and 50 viscocanalostomy) while 160 eyes had trabeculectomy.
Cillino 2005 used Mitomycin C (MMC) in all participants. El Sayyad 2000 and Russo 2008 used 5-Fluorouracil (5-FU) injections in a small proportion of both the non-penetrating filtration surgery (NPFS) and trabeculectomy arms of their studies. Kobayashi 2003 used MMC in all participants in the trabeculectomy arm. The other studies did not use an adjuvant in either arm.
Goniopunctures were used in the NPFS arm of all included studies except Russo 2008, which used reticulated hyaluronate implants in all of the deep sclerectomy operated participants and Yalvac 2004 which used hyaluronate injection in all cases (viscocanalostomy). Similarly, Kobayashi 2003 used viscocanalostomy with hyaluronate injection in all operated cases. However, their goniopuncture rate was just above half of all cases.
Laser suture lysis was only used by El Sayyad 2000.
Further details of adjuvant usage are shown in Table 1.
Types of outcome measures
All five included studies had intraocular pressure as the primary outcome. The length of follow-up was 12 months in Cillino 2005, El Sayyad 2000 and Kobayashi 2003, three years in Yalvac 2004 and four years in Russo 2008.
Secondary outcome measures
Quality of life measures
Only Kobayashi 2003, El Sayyad 2000 and Russo 2008 reported changes to medication scores ( Table 2). Whilst change in medication score clearly impacts on patients, none of the five trials used any quality of life questionnaires.
We excluded 16 studies after reviewing their full-text. The reasons for exclusion are detailed in the 'Characteristics of excluded studies' table.
Risk of bias in included studies
Methodological aspects of the included studies were generally a potential source of risk of bias. More details of methodological quality are shown in the 'Characteristics of included studies' table and in Figure 2 and Figure 3.
|Figure 2. Methodological quality graph: review authors' judgements about each methodological quality item presented as percentages across all included studies.|
|Figure 3. Risk of bias summary: review authors' judgements about each risk of bias item for each included study.|
Adequate sequence generation was noted only in Kobayashi 2003. Cillino 2005 and Russo 2008 showed inadequate sequence generation because they were quasi-randomised using surgical chart numbers while it was unclear in El Sayyad 2000 and Yalvac 2004. For three out of the five studies, we assessed allocation concealment as unclear and the remaining two studies had a high risk of bias (Cillino 2005 and Russo 2008 because the chart numbers could not be concealed).
Cillino 2005 and Russo 2008 were the only two studies where detection bias was clearly avoided by using observers masked to the intervention for the primary outcome measure. We assessed the other three studies as unclear.
Incomplete outcome data
We detected incomplete outcome data in three of the included studies with post randomisation exclusions. Kobayashi 2003 and Russo 2008 avoided this source of bias because they were paired eye studies.
We did not create funnel plots as less than 10 studies were included in the review.
Effects of interventions
The odds of success in deep sclerectomy participants was not different to that in trabeculectomy participants (odds ratio (OR) 0.98, 95% confidence interval (CI) 0.51 to 1.88) while the odds of success in viscocanalostomy participants was lower than in trabeculectomy participants (OR 0.33, 95% CI 0.13 to 0.81). We did not combine the different types of non-penetrating surgery because there was evidence of a subgroup difference when examining total success. Details and effect estimates are illustrated in Analysis 1.1; Figure 4.
|Figure 4. Forest plot of comparison: 1 Non-penetrating filtration surgery verus trabeculectomy, outcome: 1.1 Pressure control without drops.|
Similar findings were seen with success with out drops and partial success with or without drops although here we did estimate a pooled figure (OR 0.79, 95% CI 0.35 to 1.79) which was not statistically significant. Details and effect estimates are illustrated in Analysis 1.2; Figure 5.
|Figure 5. Forest plot of comparison: 1 Non-penetrating filtration surgery verus trabeculectomy, outcome: 1.2 Pressure control with or without drops.|
Summary scores for these outcomes are collated in Table 2.
Most of the participants who showed reduction in visual acuity were in the trabeculectomy group, mainly related to age-related maculopathy or cataract.
The rates of the reported adverse effects were 26 complications with non-penetrating surgeries (17%), compared to 104 complications in the trabeculectomy group (65%). Postoperative IOP spikes were reported by Cillino 2005, Kobayashi 2003 and Russo 2008 in the non-penetrating procedures group. There were no postoperative IOP spikes reported in the trabeculectomy arms of any of the included studies, except Russo 2008. Hypotony was reported more in the trabeculectomy arm than the non-penetrating procedures arm of the studies but the difference in hypotony rates differed between studies. Decrease of visual acuity due to developing cataract was reported in two participants in the viscocanalostomy group compared to seven participants in the trabeculectomy group in Yalvac 2004. Similarly, nine participants were reported to develop cataract after trabeculectomy compared to none after deep sclerectomy in Russo 2008. El Sayyad 2000 also reported one participant developing cataract in the trabeculectomy group. Kobayashi 2003 reported decrease of visual acuity from postoperative increased IOP in one participant with viscocanalostomy. Similarly in the viscocanalostomy group in Yalvac 2004, there was one participant with decrease of visual acuity from Descemet's membrane detachment. Full details of adverse effects are shown in Table 3.
Glaucoma is an important public health concern. Its irreversibility and the demographic changes of an ageing population add to the problem. The issue of intraocular pressure (IOP) control and the success rates of non-penetrating glaucoma surgery compared to trabeculectomy remain a point of ongoing debate.
Summary of main results
There was a trend toward better IOP outcomes with trabeculectomy which was statistically significant when comparing total success in participants with viscocanalostomy with trabeculectomy. Complications appeared more common in the trabeculectomy arm, where cataract was more commonly reported. None of the studies included quality of life outcome questionnaires and the methodological quality of the studies was not high. We found evidence of a subgroup difference between viscocanalostomy and deep sclerectomy when examining total success, although since this was evidence from across studies rather than within each study, it should be treated with some caution.
Overall completeness and applicability of evidence
Surgical expertise is one of the factors that can affect outcomes, and when considering two different surgical procedures, viscocanalostomy and deep sclerectomy, the difference in surgical expertise between the procedures might explain the difference in outcomes. None of the included studies commented on the surgical expertise of the surgeons performing either procedure. This should be considered when evaluating new, highly technical procedures such as non-penetrating filtration surgery (NPFS) which have been practised far less in comparison to trabeculectomy which has been widely used for over 40 years. Jonescu-Cuypers et al had first reported no success at all for viscocanalostomy but this increased to 30% in their subsequent reports (Jonescu-Cuypers 2001; Luke 2002). Similarly, Gilmour et al described how the procedure of viscocanalostomy required a learning curve and that this might have be relevant to their outcomes when compared to trabeculectomy (Gilmour 2009).
There was differences in some specific features of the interventions in the trials. It is worth mentioning that whereas viscocanalostomy groups used high-viscosity sodium hyaluronate in all eyes, only Russo 2008 used reticulated hyaluronic acid implants in all their deep sclerectomy participants, which may have modified the outcome of this study. The use of antimetabolites was not uniform among the five trials. Only Yalvac 2004 did not use antimetabolites in both groups. However, the overall rate of using operative adjuvants is more than double for the non-penetrating procedures when compared to trabeculectomy. The use of antimetabolites can directly affect the success rates of either procedure (Wilkins 2005). Russo 2008 and Yalvac 2004 did not report the use of goniopuncture which is considered by some authors as the completing step in NPFS, directly affecting success rates (Mendrinos 2008).
The justification for non-penetrating filtering surgical techniques is based on greater safety with a lower risk of complications when compared to trabeculectomy (Mendrinos 2008; Shaarawy 2004; Tan 2001). Hypotony and hyphaema were two adverse effects reported in all of the trials included in this review. Although they were recorded at lower rates in the NPFS participants than in those who had trabeculectomy, it is important to note that these risks were also low in the trabeculectomy group. Randomised controlled trials rarely have power to look reliably at adverse events and whilst we have collated all the information that we could on harms, it is important to view these data with caution.
Tan 2001 highlighted that quality of life, measured by functional status and sense of well being, is lower in patients with glaucoma compared with control participants, and is influenced by visual acuity, visual field impairment and topical medication use. In this review, little attention was paid to fields of vision in the included trials. Although both compared procedures appeared to reduce the need for medication, the difference between the two appears subtle and was only reported in three trials (El Sayyad 2000; Kobayashi 2003; Russo 2008). Visual acuity appeared to be affected mainly in the trabeculectomy group, with diversity in reporting among included trials ( Table 2). None of the five trials used quality of life measure questionnaires. This was a key finding of this review.
Quality of the evidence
We planned to report treatment effects separately for the two types of NPFS. We did not combine results for the total success comparison because there was evidence of a difference in treatment effect. Because this subgroup analysis was across studies, it should be viewed with some caution. When we combined total and qualified success rates, we still found that trabeculectomy had better outcomes compared to NPFS but these differences were not statistically significant. Larger studies would be needed to assess evidence of a treatment effect when considering this outcome. Overall, the quality of the evidence was not high. Two trials were quasi-randomised, three had post randomisation exclusions and none provided patient orientated outcomes. Surgical trials are demanding in terms of controlling bias especially for masking. Only two trials attempted to control for observation bias by masking the observers of the primary outcome.
Potential biases in the review process
Pildal 2007 highlighted that trials without adequate allocation concealment have been shown to overestimate the benefit of experimental interventions. Methodological quality issues were a strong source of bias in most of the included trials. With stricter methodological inclusion criteria, none of the five trials would have been included. It is important to consider avoiding these sources of bias in future trials.
Agreements and disagreements with other studies or reviews
The meta-analysis by Cheng et al focused on the pooled success rates of viscocanalostomy and deep sclerectomy rather than comparative studies with specified qualifying criteria (Cheng 2004). In this review the primary outcome measure is similar to that described by Chen 1997: surgical total success when IOP is less than 21 mmHg without additional medications after one year of surgery. They reported that the probability of successful control of IOP was 82% at five years and 67% at 10 and 15 years. Ke 2011 conducted a meta-analysis, in which their study inclusion criteria included all non-penetrating trabecular surgeries as one entity. Similar to our review, they concluded that trabeculectomy could reduce IOP better than non-penetrating trabecular surgeries which, however, showed lower rates of complications when compared to trabeculectomy.
Implications for practice
This review provides limited evidence that control of intraocular pressure (IOP) is better with trabeculectomy than viscocanalostomy although there is greater uncertainty around the effect with deep sclerectomy. The confidence limits are wide and the quality of evidence poor so one cannot conclude this might indicate equivalence. Results regarding harms were inconclusive but this is not surprising given that adverse events are often rare. This review has highlighted the lack of use of quality of life outcomes and the need for randomised controlled trials (RCTs) with higher methodological quality to address these issues.
Implications for research
A high-quality, multi-centred RCT is required to compare trabeculectomy to either deep sclerectomy or viscocanalostomy. We feel that these techniques should not be combined in one group when compared to trabeculectomy. Surgical expertise should be taken into account when allocating centres for a large RCT, i.e. the surgeons should be undertaking a defined minimum number of procedures in a year. Alternatively, an expert design could be used where participants are randomised to "expert surgeons" for either technique. Since it is unlikely that better IOP control will be offered by non-penetrating filtration surgery but that these techniques offer potential gains for patients in terms of quality of life, we feel that the trial should be a non-inferiority design with quality of life measures. Complications should be well defined with rigorous reporting standards and methods.
We greatly appreciate the work of Dr. Nitin Anand, Dr. Nahla Sobhy and Dr. Mohamed Khafagy for their help in the review preparation. We acknowledge the support of Anupa Shah in the preparation of the review, together with all the editorial team of the Cochrane Eyes And Vision Group (CEVG). We thank Jennifer Burr, Ann Ervin and Kristina Lindsley for their comments on the protocol and Marie Diener-West, Ruth Thomas and Scott Fraser for their comments on the review. CEVG created and executed the electronic search strategies.
Data and analyses
- Top of page
- Authors' conclusions
- Data and analyses
- Contributions of authors
- Declarations of interest
- Sources of support
- Differences between protocol and review
- Index terms
Appendix 1. CENTRAL search strategy
#1 MeSH descriptor Glaucoma, Open-Angle
#2 open near angle near glaucoma*
#4 primary near glaucoma*
#5 chronic near glaucoma*
#6 secondary near glaucoma*
#7 low near tension near glaucoma*
#8 low near pressure near glaucoma*
#9 normal near tension near glaucoma*
#10 normal near pressure near glaucoma*
#11 pigment near glaucoma*
#12 MeSH descriptor Exfoliation Syndrome
#13 exfoliat* near glaucoma*
#14 pseudoexfoliat* near syndrome*
#15 pseudoexfoliat* near glaucoma*
#16 (#1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15)
#17 MeSH descriptor Trabeculectomy
#19 MeSH descriptor Sclerostomy
#23 MeSH descriptor Filtering Surgery
#24 filtrat* near surg*
#25 (#17 OR #18)
#26 (#19 OR #20 OR #21 OR #22 OR #23 OR #24)
#27 (#25 AND #26)
#28 (#16 AND #27)
Appendix 2. MEDLINE (OvidSP) search strategy
1. randomized controlled trial.pt.
2. (randomized or randomised).ab,ti.
9. exp animals/
10. exp humans/
11. 9 not (9 and 10)
12. 8 not 11
13. exp glaucoma open angle/
14. (simple$ adj3 glaucoma$).tw.
15. (open adj2 angle adj2 glaucoma$).tw.
17. (primary adj2 glaucoma$).tw.
18. (chronic adj2 glaucoma$).tw.
19. (secondary adj2 glaucoma$).tw.
20. (low adj2 tension adj2 glaucoma$).tw.
21. (low adj2 pressure adj2 glaucoma$).tw.
22. (normal adj2 tension adj2 glaucoma$).tw.
23. (normal adj2 pressure adj2 glaucoma$).tw.
24. (pigment$ adj2 glaucoma$).tw.
25. exp exfoliation syndrome/
26. (exfoliat$ adj2 syndrome$).tw.
27. (exfoliat$ adj2 glaucoma$).tw.
28. (pseudoexfoliat$ adj2 syndrome$).tw.
29. (pseudoexfoliat$ adj2 glaucoma$).tw.
31. exp trabeculectomy/
34. exp sclerostomy/
38. exp filtering surgery/
39. (filtrat$ adj3 surg$).tw.
41. 33 and 40
42. 30 and 41
43. 12 and 42
The search filter for trials at the beginning of the MEDLINE strategy is from the published paper by Glanville (Glanville 2006).
Appendix 3. EMBASE (OvidSP) search strategy
1. exp randomized controlled trial/
2. exp randomization/
3. exp double blind procedure/
4. exp single blind procedure/
7. (animal or animal experiment).sh.
9. 7 and 8
10. 7 not 9
11. 6 not 10
12. exp clinical trial/
13. (clin$ adj3 trial$).tw.
14. ((singl$ or doubl$ or trebl$ or tripl$) adj3 (blind$ or mask$)).tw.
15. exp placebo/
18. exp experimental design/
19. exp crossover procedure/
20. exp control group/
21. exp latin square design/
23. 22 not 10
24. 23 not 11
25. exp comparative study/
26. exp evaluation/
27. exp prospective study/
28. (control$ or prospectiv$ or volunteer$).tw.
30. 29 not 10
31. 30 not (11 or 23)
32. 11 or 24 or 31
33. exp open angle glaucoma/
34. (open adj2 angle adj2 glaucoma$).tw.
36. (primary adj2 glaucoma$).tw.
37. (chronic adj2 glaucoma$).tw.
38. (secondary adj2 glaucoma$).tw.
39. (low adj2 tension adj2 glaucoma$).tw.
40. (low adj2 pressure adj2 glaucoma$).tw.
41. (normal adj2 tension adj2 glaucoma$).tw.
42. (normal adj2 pressure adj2 glaucoma$).tw.
43. (pigment$ adj2 glaucoma$).tw.
44. exp exfoliation syndrome/
45. (exfoliat$ adj2 syndrome$).tw.
46. (exfoliat$ adj2 glaucoma$).tw.
47. (pseudoexfoliat$ adj2 syndrome$).tw.
48. (pseudoexfoliat$ adj2 glaucoma$).tw.
50. exp trabeculectomy/
53. exp glaucoma surgery/
57. exp filtering operation/
58. (filtrat$ adj3 surg$).tw.
60. 52 and 59
61. 49 and 60
62. 32 and 61
Appendix 4. LILACS search strategy
glaucoma$ and open or chronic or primary or low or normal or pigmentary or exfoliat$ and trabeculectom$ or sclerostom$ or sclerectom$ or viscocanalostom$
Appendix 5. metaRegister of Controlled Trials search strategy
(trabeculectomy) and (sclerostomy or sclerectomy or viscocanalostomy)
Appendix 6. ClinicalTrials.gov search strategy
(Trabeculectomy) AND (Sclerostomy OR Sclerectomy OR Viscocanalostomy)
Appendix 7. ICTRP search strategy
Trabeculectomy = Condition AND Sclerostomy OR Sclerectomy OR Viscocanalostomy = Intervention
Protocol first published: Issue 2, 2008
Review first published: Issue 2, 2014
Contributions of authors
Conceiving the review: ME
Designing the review: ME
Co-ordinating the review: ME
Data collection for the review
- Designing electronic search strategies: CEVG Trials Search Co-ordinator
- Undertaking manual searches: ME, MK
- Screening search results: ME, MK, OE
- Organising retrieval of papers: ME
- Screening retrieved papers against inclusion criteria: ME, MK
- Appraising quality of papers: ME, MK
- Extracting data from papers: ME, MK
- Writing to authors of papers for additional information: ME, MK
- Providing additional data about papers: ME, MK
- Obtaining and screening data on unpublished studies: MK
Data management for the review
- Entering data into RevMan: ME, CB, MK
- Analysis of data: ME, CB, MK, OE
Interpretation of data
- Providing a methodological perspective: ME, MK, CB
- Providing a clinical perspective: ME, MK, OE
- Providing a policy perspective: ME, MK, OE
- Providing a consumer perspective: ME, MK, OE
Writing the review: ME, MK, OE, RW
Providing general advice on the review: ME, CB, MK, OE
Securing funding for the review: ME
Declarations of interest
Sources of support
- NIHR, UK.CB and RW acknowledge financial support from the Department of Health through the award made by the National Institute for Health Research to Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology for a Specialist Biomedical Research Centre for Ophthalmology. The views expressed in this publication are those of the authors and not necessarily those of the Department of Health
- No sources of support supplied
Differences between protocol and review
In the 'Assessment of risk of bias in the included studies' section two parameters to assess the risk of bias were added in the review that were not in the protocol: adequate sequence generation in selection bias and selective reporting of outcomes.
Medical Subject Headings (MeSH)
MeSH check words
Aged; Humans; Middle Aged
* Indicates the major publication for the study