Figure 1. Causal Pathway of Treatment of Coexisting Cataract & Glaucoma
 |
|
|
The Agency for Healthcare Research and Quality (AHRQ), through its Evidence-Based Practice Centers (EPCs), sponsors the development of evidence reports and technology assessments to assist public- and private-sector organizations in their efforts to improve the quality of health care in the United States. The reports and assessments provide organizations with comprehensive, science-based information on common, costly medical conditions and new health care technologies. The EPCs systematically review the relevant scientific literature on topics assigned to them by AHRQ and conduct additional analyses when appropriate prior to developing their reports and assessments.
To bring the broadest range of experts into the development of evidence reports and health technology assessments, AHRQ encourages the EPCs to form partnerships and enter into collaborations with other medical and research organizations. The EPCs work with these partner organizations to ensure that the evidence reports and technology assessments they produce will become building blocks for health care quality improvement projects throughout the Nation. The reports undergo peer review prior to their release.
AHRQ expects that the EPC evidence reports and technology assessments will inform individual health plans, providers, and purchasers as well as the health care system as a whole by providing important information to help improve health care quality.
We welcome written comments on this evidence report. They may be sent to: Director, Center for Practice and Technology Assessment, Agency for Healthcare Research and Quality, 6010 Executive Blvd., Suite 300, Rockville, MD 20852.
Carolyn M. Clancy, M.D.
Director
Agency for Healthcare Research and Quality
Jean Slutsky, P.A., M.S.P.H.
Acting Director, Center for Practice and Technology Assessment
Agency for Healthcare Research and Quality
The authors of this report are responsible for its content. Statements in the report should not be construed as endorsement by the Agency for Healthcare Research and Quality or the U.S. Department of Health and Human Services of a particular drug, device, test, treatment, or other clinical service.
Objectives. Cataract and glaucoma are ocular diseases that often coexist, with prevalences over age 40 of 20 percent and two percent, respectively. There is no agreement concerning their optimal management when coexistent. We prepared this evidence report to: identify the important questions pertinent to surgical treatment of coexisting cataract and glaucoma; assess the quality and content of evidence on surgical treatment of coexisting cataract and glaucoma; and inform clinical practitioners and identify areas for future research.
Search Strategy. The searches were conducted on publications from 1980 to April 2000 using two electronic databases, PubMed and CENTRAL. Text words and Medical Subject Heading (MeSH) search terms used included (“cataract” and “glaucoma”) and (“surgery” or “filtering surgery” or “cataract extraction” or “sclerostomy” or “trabeculectomy” or “phacoemulsification”). We included variants of “randomized controlled trials” in the search as well as “case report” and “case series”. The electronic searches were augmented by a hand search of primary journals.
Selection Criteria. Two individuals independently reviewed each abstract. The exclusion criteria were: no human data, lack of adults, lack of original data, failure to address open-angle or primary angle- closure glaucoma, not a controlled trial nor a case series greater than or equal to 100 eyes, only addressing intracapsular surgery, only addressing full-thickness procedures, was a meeting abstract, and not in English.
Data Collection and Analysis. Two data abstraction forms were developed. The 25 question quality assessment form was divided into the following categories: representativeness, bias and confounding, description of therapy, outcomes and followup, and statistical quality and interpretation. A content assessment form was developed through an iterative process.
Each article was reviewed by two reviewers, at least one trained in research methodology and at least one trained in ophthalmology. Quality scores for each controlled trial and cohort study were tabulated.
For each study question one member of the analysis team summarized the extracted data and formulated a conclusion about the answer to each question. For each conclusion, the entire study team assigned an evidence grade of A (strong), B (intermediate), C (weak), or I (insufficient).
Main Results. There was strong evidence that glaucoma surgery is associated with an increased risk of postoperative cataract; moderately strong evidence that mitomycin-C, but not 5-fluorouracil is beneficial in combined procedures, limbus- and fornix-based conjunctival incisions are equally effective for lowering IOP, and the size of the phacoemulsification incision is not important; and weak evidence that combined procedures using phacoemulsification rather than nuclear expression result in lower long-term IOP, as do two-site compared to one-site combined procedures.
Limitations of the literature included lack of optic nerve and visual field data, lack of objective description of the ocular lens, inconclusive information on complications, lack of patient preference and quality of life data, and limited followup in many studies.
Conclusions. The literature does not point to one optimal strategy for controlling IOP in patients with coexisting cataract and glaucoma needing surgery. Therefore, there is a continued need for high quality studies with greater duration and more information on optic nerve and visual field findings.
Cataract and glaucoma are both common conditions and are often present in the same patient. Cataract surgery is the most commonly performed surgical procedure on Medicare beneficiaries. In 1998, approximately 1.4 million cataract surgeries were performed on Medicare beneficiaries. Primary open-angle glaucoma affects at least 2.5 million individuals in the United States, predominantly adults over 50 years of age. The total direct cost expenditures for glaucoma therapy have been estimated at $1.56 billion per year.
Although guidelines exist for the indications for cataract surgery in the otherwise healthy eye and for glaucoma surgery in eyes with glaucoma, there is controversy concerning the indications for surgery when both cataract and glaucoma are present. In addition, there is no clear consensus about the appropriate timing of the surgery for either condition, or about the best surgical technique. Furthermore, there is no agreement concerning the optimal management of these disorders when coexistent. Therefore, the Agency for Healthcare Research and Quality (AHRQ), (formerly the Agency for Health Care Policy and Research), awarded a contract to the Johns Hopkins University Evidence-based Practice Center to prepare an evidence report on the topic. This evidence report was undertaken to identify the most important questions pertinent to surgical treatment of coexisting cataract and glaucoma, to assess the quality and content of the evidence on surgical treatment of coexistent cataract and glaucoma, and to inform clinical practitioners and identify areas where future research is needed.
The target population addressed in this evidence report is adult patients with coexisting cataract and glaucoma. Both open- and closed-angle glaucoma were considered. The target audience for the report is health care professionals involved in the care of adult patients with coexisting cataract and glaucoma.
The following questions were addressed in the evidence report:
What are the risks and benefits of staged versus combined procedures for coexisting cataract and glaucoma?
What is the effect of cataract surgery on short- and long-term intraocular pressure (IOP) control in open-angle glaucoma patients?
What is the effect of glaucoma surgery (i.e., filtration surgery) on the development and progression of cataract?
What are the risks and benefits of antifibrosis agents in the surgical treatment of coexisting cataract and glaucoma?
What are the risks and benefits of trabeculectomy versus endoscopic laser versus deep sclerectomy/viscocanulostomy in patients with coexisting cataract and glaucoma?
What are the risks and benefits of single-site versus two-site operations for coexisting cataract and glaucoma?
What are the risks and benefits of nuclear expression or phacoemulsification in patients with coexisting cataract and glaucoma?
The interventions evaluated included two commonly used methods for cataract extraction: nuclear expression and phacoemulsification. Several methods used for the surgical management of glaucoma were studied, including laser treatment, filtration surgery, and the newer techniques of endoscopic cyclophotocoagulation and the so-called nonpenetrating procedures. The use of the antifibrosis agents 5-fluorouracil and mitomycin C was also evaluated. The various ways in which cataract and glaucoma surgery can be performed either simultaneously or sequentially were studied.
The primary outcome assessed for the questions involving glaucoma surgery was IOP, both in the immediate postoperative period and long term. The articles reviewed generally lacked useful data on optic nerve appearance and visual fields, thereby precluding a meaningful review of those outcomes. Likewise, the lack of information concerning complications limited the ability to address the risk aspects of each question. For the question concerning the development of cataract after glaucoma surgery, both the observation of a cataract and the performance of cataract surgery were assessed outcomes.
The literature searches were conducted using two electronic databases, PubMed and CENTRAL, the Cochrane Collaboration's database. The textwords and medical subject heading (MeSH) terms we used included (“cataract” and “glaucoma”) and (“surgery” or “filtering surgery” or “cataract extraction” or “sclerostomy” or “trabeculectomy” or “phacoemulsification”). To retrieve the appropriate types of studies, we included variants of randomized controlled trials in the search as well as “case report” and “case series.” We limited the search to publications in English. Citations spanned the period from 1980 to April 2000. The electronic searches were augmented by a hand search of primary journals.
The abstracts of all identified citations were independently reviewed by two members of the study team to determine if they met eligibility criteria. When reviewing abstracts, the following criteria were used to exclude articles from further consideration: lack of adults in the study population, lack of original data, failure to include human data, failure to address open-angle glaucoma or primary angle closure glaucoma, neither a controlled trial nor a case series greater than or equal to 100 eyes, addressing only intracapsular cataract surgery, and addressing only full-thickness procedures. Meeting abstracts and texts not in English were also excluded. Disagreements between reviewers about the eligibility of a citation were adjudicated at a meeting of the entire study team. We reviewed the full text of the article for a 5-percent random sample of rejected abstracts. None of these articles was deemed eligible for inclusion in the full article review.
Two forms were developed and used to review eligible citations. The quality assessment form contained questions grouped into the following categories:
Representativeness of study population—how completely the authors described the study subjects.
Bias and confounding, including completeness of randomization and masking.
Description of therapy—completeness of the description of the study protocol and other treatments, if applicable. Outcomes and followup, including explicit description of the outcomes reported (e.g., assessment of IOP, cataract assessment) and length of followup.
Statistical quality and interpretation.
Quality scores for each controlled trial and cohort study were tabulated.
The article content assessment form was developed through an iterative process. The study team discussed which outcomes should be included on the form. Following these discussions, selected relevant articles were reviewed to create a standardized abstraction of reported outcomes in a manner that was applicable across the range of studies. Finally, the form was pilot tested on a sample of articles and revised as necessary. Instructions for the use of the form were developed by consensus for the purposes of consistency.
Each article identified by the search process and determined to meet eligibility criteria through the abstract process was reviewed by two reviewers. At least one of the reviewers was trained in research methodology and at least one was trained in ophthalmology or internal medicine.
Data from the article review process were entered into a relational database. For each study question, one member of the analysis team summarized the available data and reached a conclusion about the answer to the question. The entire study team assigned an evidence grade to the conclusions reached about each study question.
Of 919 unique, potentially relevant citations, 131 articles were eligible for full article review. After 21 were excluded, 110 articles remained for data extraction. Study quality scores were calculated for controlled trials and cohort studies, comprising 81 articles. The mean overall study quality score was 53 percent.
The findings are as follows:
Glaucoma surgery was associated with an increased risk of postoperative cataract.
A glaucoma procedure added to cataract surgery lowers IOP more than cataract surgery alone.
Limbus- and fornix-based conjunctival incisions provided the same degree of long-term IOP lowering in combined surgery.
In combined surgery using phacoemulsification, the size of the cataract incision did not affect long-term IOP control.
5-fluorouracil was not beneficial in further lowering IOP when used with combined procedures.
Mitomycin C was efficacious in producing lower long-term IOPs when used with combined procedures.
Combined procedures resulted in lower IOP at 24 hours than cataract extraction alone.
Extracapsular cataract extraction (ECCE) alone appears to increase IOP at 24 hours.
In the long term, cataract surgery alone lowered IOP by 2 to 4 mm Hg, combined cataract and glaucoma surgery lowered IOP by 6 to 8 mm Hg, and the performance of a glaucoma procedure alone provided even greater long-term IOP lowering than combined cataract and glaucoma surgery.
Combined surgery in which the incisions for the cataract extraction and glaucoma procedure are separate provided slightly lower long-term IOP than a one-site approach.
Combined surgery in which phacoemulsification is used provided slightly lower long-term IOP than nuclear expression.
The same complications occurred with combined surgery as occur with cataract and glaucoma surgery performed separately.
5-fluorouracil use was associated with epithelial defects.
Hypotony was more likely after the use of antifibrosis agents than after surgery performed without antifibrosis agents.
The literature addressing management of coexistent cataract and glaucoma has important limitations that future research should consider and address. First, those limitations pertaining to the entire literature are described, and then those that are specific to individual questions are described.
In terms of the management of glaucoma, because the extant literature concentrates on IOP, this review was limited to the effect of surgery on IOP. Future studies should assess the effect of surgery on the appearance of the optic nerve, visual field, and quality of life. Because of the chronic and slowly progressive nature of glaucoma, these parameters need to be evaluated in long-term (5 years or more) studies.
In terms of cataract, future studies should pay more attention to the quantification of cataract and its effect on vision. The existing literature is generally lacking in any objective description of the ocular lens, whether clear or cataractous. Quantifying lens opacity and assessing its effect on vision and vision-related quality of life should be a priority in future studies. Standardized grading systems for cataract (already in use) should be employed.
Although the literature represented populations in North America, Europe, and Asia, few of the reports included significant numbers of black Americans. This population has a high prevalence of glaucoma and may respond differently to glaucoma surgery. Future researchers should strive to include more black Americans in their studies.
Assessment of the evidence was limited by the quality of the literature. In several of the study quality categories, particularly bias and confounding, there was tremendous room for improvement. A lack of comparability of study groups and an absence of masking of observers greatly compromised the validity of many studies. In future studies, it would be valuable to strive for comparability of study groups and statistical adjustment for baseline differences, as well as more objective measures for assessing outcomes.
The following specific questions are areas for future research, both because of their importance and the fact that questions remain unresolved by the extant literature:
Does trabeculectomy alone lower IOP more than phacoemulsification plus trabeculectomy?
Does phacoemulsification harm functioning filtering blebs?
Does phacoemulsification reliably result in lower IOP on the first day after surgery? If so, then it might not be necessary to perform combined surgery.
Does a two-stage procedure provide lower IOP than a combined procedure? What is the role of mitomycin C in combined procedures and how should it be administered?
Is the finding of decreased posterior capsular opacification with the use of mitomycin C reproducible?
Can better glaucoma procedures be developed to be used in conjunction with cataract surgery?
What are the risk factors for cataract development and progression following glaucoma surgery?
Cataract and glaucoma are both common conditions and are often present in the same patient. Cataract surgery is the most commonly performed surgical procedure on Medicare beneficiaries.1 Approximately 1.6 million cataract surgeries were performed on Medicare beneficiaries in 1998.a Primary open-angle glaucoma affects at least 2.5 million individuals in the United States.2 The total direct cost expenditures for glaucoma therapy have been estimated at $1.56 billion per year.3 Although guidelines exist for the indications for cataract surgery in the otherwise healthy eye and for glaucoma surgery in eyes with glaucoma, controversy exists concerning the indications for surgery when both cataract and glaucoma are present. In addition, there is no clear consensus about the appropriate timing of the surgery for each condition, or about the best surgical technique. At the present time, there is no agreement concerning the optimal management of these disorders when coexistent.
The interventions considered in this report are the surgical approaches to the treatment of glaucoma and cataract. For the former, these approaches comprise laser treatment, filtration surgery, and the newer techniques of endoscopic cyclophotocoagulation and the so-called nonpenetrating operations. For cataract surgery, the approaches are corneal and scleral extracapsular extraction using either phacoemulsification or nuclear expression. This report also considers specific medications (i.e., mitomycin-C and 5-fluorouracil) that may be used as ancillary therapy in conjunction with surgical treatment of glaucoma. However, it does not examine the role of glaucoma medications in the treatment of coexisting glaucoma and cataract because our clinical experts felt that it was most important to focus on the surgical treatment options (see the section on Identifying the Specific Questions in Chapter 2).
The affected population is primarily adults, with a predominance of patients over 50 years of age. The estimated size of the affected population in the United States is 100,000 to 150,000 per year. The usual intervention setting is outpatient surgery in a hospital or in an ambulatory surgery center.
The primary outcome for glaucoma is intraocular pressure (IOP), because the reviewed studies were generally of insufficient duration to draw conclusions about optic nerve and visual field changes. Furthermore, few studies reported optic nerve or visual field findings in a manner permitting meaningful analysis.
The objectives of this evidence report are to: identify the most important questions pertinent to treatment of patients with coexisting cataract and glaucoma; assess the published literature with respect to quality and content regarding these questions; and to inform clinical practitioners and identify areas where future research is needed, based on the literature findings.
We identified a core group of four clinically oriented technical experts who provided extensive input throughout the project. The group included two representatives from our principal partner, the American Academy of Ophthalmology (AAO). The core group of technical experts were: Paul Lee, MD, JD; Stephen Obstbaum, MD; Joseph Caprioli, MD (AAO representative); and Ronald L. Gross, MD (AAO representative).
Representatives from a range of other stakeholder organizations including the American Glaucoma Society, the American Optometric Association, the American Society of Cataract and Refractive Surgery, the Eyes and Vision Group of the Cochrane Collaboration, the Glaucoma Research Foundation, the Health Care Financing Administration, and The Glaucoma Foundation were identified as peer reviewers. The representatives included two glaucoma patients. A list of these peer reviewer representatives is attached as Appendix A.
The target population addressed by this evidence report is adult patients (age 18 and older) with coexisting cataract and glaucoma. Both open-angle and primary closed-angle glaucoma were considered. The target audience is health care professionals involved in the care of adult patients with coexisting cataract and glaucoma.
The study team devised an initial list of potential questions. The preliminary list of questions was discussed with ophthalmologists external to the study team to add other relevant questions. Following these discussions, the study investigators prepared a letter which included a list of the potential questions regarding the treatment of coexisting cataract and glaucoma (Appendix B). Through the use of this letter, each technical expert and peer reviewer was asked to rate the importance of each proposed question from the perspective of his/her own profession or organization. Based on the responses to the letter and a conference call with the technical experts, consensus was reached regarding the final questions to be addressed by this evidence report. The following questions are addressed in the evidence report:
What are the risks and benefits of staged versus combined procedures for coexisting cataract and glaucoma?
What is the effect of cataract surgery on short- and long-term intraocular pressure control (IOP) in open-angle glaucoma patients?
What is the effect of glaucoma surgery (i.e., filtration surgery) upon the development and progression of cataract?
What are the risks and benefits of anti-fibrosis agents in combined cataract and glaucoma procedures?
What are the risks and benefits of trabeculectomy versus endoscopic laser versus deep sclerectomy/viscocanulostomy for coexistent cataract and glaucoma?
What are the risks and benefits of single-versus two-site operations for coexistent cataract and glaucoma?
What are the risks and benefits of nuclear expression or phacoemulsification in patients with glaucoma and cataract?
Studies reporting short-term outcomes were defined as those whose reported followup was 24 hours. Studies reporting long-term outcomes were defined as those reporting followup of 6 months or greater in length.
A model of the causal pathway for treatment of coexisting cataract and glaucoma was constructed (Figure 1
.). Patient outcomes are influenced by all the factors shown. Provider factors such as subspeciality and geographic location, patient factors such as age and severity of ocular disease, and health system factors all influence the management strategy chosen for treating coexisting cataract and glaucoma. In turn, these factors all influence patient outcomes.A number of literature sources were used in order to identify all studies that were potentially relevant to the study questions examined in this evidence report. Preliminary searches were completed on PubMed, the web-based access to MEDLINE provided by the National Library of Medicine. The results of the searches were used in refinement of the questions and formed the basis for the database of citations used for this evidence report.
The searches were conducted using two electronic databases, PubMed and CENTRAL, the Cochrane Collaboration's database. The electronic searches were augmented by a hand search of five ophthalmology journals identified by the principal investigators as high priority (Appendix C). The hand search included articles published through April 2000.
A core search strategy was developed and was used to search both the CENTRAL and PubMed databases. For the PubMed searches, the core search strategy was combined with an optimal search strategy for the identification of controlled clinical trials.4 Text words and the Medical Subject Heading (MeSH) term “case report” were added to the search strategy to allow for the inclusion of case studies. Because case studies are a source of information on reported complications of surgery, we felt it was important to include them in the literature review. Multiple searches were completed. The first search of PubMed was completed November 22, 1999. The strategies used in the searches are included in Appendix D.
The results of all searches were downloaded or hand entered into a database of studies that were considered potentially relevant to the literature review. The abstracts of all identified citations were independently reviewed by two members of the study team to determine if they met eligibility criteria. An abstract review form was developed based on the abstract review form used in the Evidence Report on Management of New Onset of Atrial Fibrillation5 and the Evidence Report on Anesthesia Management During Cataract Surgery.6 A copy of the abstract review form may be found in Appendix E. All the citations not meeting the eligibility criteria were coded with a reason for exclusion. For those articles without an abstract or where the reviewer could not determine whether the abstract met the eligibility criteria, a full copy of the article was obtained and reviewed. Disagreements between reviewers about eligibility were adjudicated at a meeting of the study team. A database, developed using ProCite software, was used to facilitate the tracking of search results, the abstract review process, and the printing of reports regarding the relevant literature. It is possible that the contents of an abstract may not accurately convey the findings of the published paper. In order to insure that pertinent articles were not wrongly excluded based on review of the abstract, we reviewed the full article for a five percent sample of rejected abstracts. The methods used for this review were the same as for the abstract review. Results of this review confirmed that all (100 percent) were not eligible for inclusion in the full article review.
The following criteria were applied at the abstract review phase to exclude articles from further consideration:
did not include human data
did not include adults as part of study population
had no original data
did not address open-angle glaucoma or primary angle closure glaucoma
was neither a controlled trial nor a case series greater than or equal to 100 eyes
only addressed intracapsular cataract extraction
only addressed full thickness glaucoma surgery
was a meeting abstract
was not in English.
We chose to exclude case series with less than 100 eyes because we felt they were unlikely to include sufficient data about complications. We did not include foreign language articles because we did not have the resources or the personnel to search, retrieve, and translate them.
The abstract review was also used to identify which of the following study topics each article addressed: development of cataract after glaucoma surgery; acute control of IOP after cataract surgery in glaucoma patients; long-term control of IOP after cataract surgery in glaucoma patients; and preferred surgical technique.
Each article identified by the search process, and determined to meet eligibility criteria through the abstract review process, was reviewed by two reviewers. At least one of the reviewers was an individual trained in research methodology and at least one (and usually both) trained in ophthalmology or internal medicine. In addition to Drs. Bass, Friedman, Jampel, and Lubomski, four ophthalmologists assisted in the review process: Dr. Nathan Congdon, M.D., M.P.H, an ophthalmologist trained in study methodology; Dr. Hana Levkovitch-Verbin, M.D., an ophthalmologist; Dr. John Kempen, M.D., M.P.H., an ophthalmologist trained in study methodology; and Dr. Harry Quigley, M.D., an ophthalmologist.
Two data abstraction forms were developed and extensively pilot tested and revised prior to their use. The quality assessment form is included as Appendix F and the article content assessment form is included as Appendix G.
The quality assessment form included the same exclusion criteria included on the abstract review form, and questions designed to assess the quality of the study. The form was developed based on the form used in the Evidence Report on Anesthesia Management During Cataract Surgery.6 The form contained 25 questions which where grouped into the following categories: representativeness of study population - how completely the authors described the study subjects; bias and confounding - including completeness of randomization and masking; description of therapy - completeness of the description of the study protocol and other treatments (if applicable); outcomes and followup - which includes explicit description of the outcomes reported (e.g., assessment of IOP, cataract assessment) and length of followup; and statistical quality and interpretation. Each response to a question was assigned a value from zero to two. The score for each category of study quality is the percentage of the total points available in each category. The overall quality score is the average of its five categorical scores. Quality assessment was completed for controlled trials and cohort studies only.
The article content assessment form was developed through an iterative process. The study team discussed which outcomes should be included on the form. Following these discussions, selected relevant articles were reviewed to create a standardized abstraction of reported outcomes in a manner that was applicable across the range of studies. Finally, the form was pilot tested on sample manuscripts and revised as necessary. Instructions for the use of the form were developed by consensus for the purpose of consistency. The article content assessment form was completed for all articles meeting eligibility criteria.
The article review was conducted serially. Each article was assigned a primary reviewer. The primary reviewer performed the data abstraction and completed the data abstraction forms. The secondary reviewer, using the data abstraction forms completed by the primary reviewer, reviewed the data abstraction. Differences between the reviewers were reconciled by the two abstracters in a conference to reach consensus. If a consensus could not be reached, the differences were discussed and reconciled at a meeting of all of the study investigators. The reviewers were not masked with regard to author, institution, or journal.
A relational database was developed to collect, maintain, and analyze the data abstracted during the article review phase. This database was subsequently used for construction of evidence tables.
This table includes information on the source of the article, study design, main study aim, sample size (number of eyes), proportion of African-American patients, mean patient age, principal outcomes reported, mean length of study followup, study limitations, overall quality score, and a concluding statement of each study's results.
This table includes information on the source of the article, the study treatment(s), location of conjunctival incision, description of cataract surgery, description of trabeculoplasty, description of endoscopic laser, concentration and duration of mitomycin-C administered, concentration and duration of intraoperative 5-fluorouracil administered, and dose and number of injections of postoperative 5-fluorouracil received.
This table includes information on the source of the article, exclusion criteria, scores on the study quality categories, and the overall study quality score. Scores have been calculated for randomized controlled trials, non-randomized controlled trials, and cohort studies only. Studies of case-control or clinical series design are therefore not listed in this table.
Evidence grades were assigned to statements comparing alternative techniques (e.g., IOP control using mitomycin-C agents versus no mitomycin-C, or IOP control following combined phacoemulsification/trabeculectomy versus trabeculectomy alone). The summary for each relevant section ends with a “conclusion” statement and the evidence grade. The evidence underlying the concluding statement was independently reviewed by Drs. Bass, Congdon, Friedman, Jampel, Kempen, Lubomski, and Quigley. The method used to grade the evidence is based on the method used in the Evidence Report on Anesthesia Management During Cataract Surgery.6 That scheme was based on a modification of the grading scheme employed by Garbutt and colleagues.7 The following strategy was used to grade the strength of the evidence.
Grade A (strong): Appropriate data are available for evaluating efficacy/safety of the interventions in question, including at least one well done randomized controlled trial; overall, the population of patients studied is sufficiently large, and adequate controls have been used; data are consistent and indicate that the efficacy/safety of the intervention have been accurately described and that one intervention is clearly superior, equivalent, or inferior to another in terms of one or more well-defined outcomes.
Grade B (moderate): Appropriate data are available for evaluating safety/efficacy of the interventions in question; overall, the population of patients studied is sufficiently large and adequate controls have been used; data are reasonably but not entirely consistent and indicate that the efficacy/safety of the intervention have been accurately described; the data indicate superiority (or equivalence) of one intervention over another for specific, well-defined outcomes; however, there is insufficient evidence to conclude definitively that the efficacy/safety of one intervention are clearly superior or inferior to another.
Grade C (weak): Some data are available for evaluating safety/efficacy of the interventions in question; overall, the population of patients studied is reasonably large; data indicate a trend supporting a benefit (or equivalence) of one intervention over another for specific, well-defined outcomes; however, there is insufficient evidence to conclude that one intervention is truly superior (or equivalent to) to another.
Grade I (insufficient): Appropriate data are not available or there is an insufficient number of patients studied to assess safety/efficacy of the intervention in question alone or in comparison to alternatives.
Disagreements between reviewers regarding the evidence grade were discussed and reconciled at a meeting of all of the participants in the evidence grading process. The initial evidence grades assigned by the participants may be found in Appendix I. The grades resulting from the adjudication process may be found at the conclusion of the relevant section of Chapter 3.
Based upon our completed review of the literature (as described in Chapter 3: Results) the scientific assumptions underlying the justification for a meta-analysis were not met. Most importantly, studies were sufficiently heterogeneous with regard to study aims, content, design, and the way in which the outcomes were measured, to prevent pooling of the data in a meaningful or acceptable fashion.
No supplemental analyses were completed as a part of this project.
The literature search identified 919 unique, potentially relevant citations to be included in the abstract review. A summary of the searches completed, strategy used, date of search, number of citations retrieved, and number of unique citations is included as Appendix J.
All 919 eligible abstracts were independently reviewed by two members of the study team to determine if they met eligibility criteria. Disagreements between reviewers about whether to include an abstract or citation were adjudicated at a meeting of the entire study team. The last electronic search was completed July 18, 2000, and the hand searching and abstract review process ended April 15, 2000. The abstract review identified 131 citations as eligible for full article review. Abstracts were deemed ineligible for full article review if they were not applicable to the study questions (1.8 percent of excluded abstracts), if the citation met an exclusion criterion (98 percent of excluded abstracts), or for a combination of the above. A summary of the reasons for exclusion based on the abstract review may be found in Appendix K.
At the abstract review level, the citations were tagged as referring to four study topics: development of cataract after glaucoma surgery; acute control of intraocular pressure (IOP) after cataract surgery in glaucoma patients; long-term control of IOP after cataract surgery in glaucoma patients; and preferred surgical technique (including assessment of staged versus combined procedures, efficacy of anti-fibrosis agents in surgical treatment, one- versus two-site surgery, trabeculectomy versus endoscopic laser treatment, and nuclear expression versus phacoemulsification). The majority of the abstracts (57 percent) were classified as addressing preferred surgical technique (See Appendix L for a summary of topics.)
The 131 citations deemed eligible for full article review were reviewed to assess study quality and to extract relevant data. Of the 131 articles reviewed, 21 (16 percent) were excluded during the full article review (Appendix M), leaving a total of 110 (84 percent) articles for the data extraction. The reasons for exclusion and number of articles excluded, by reason for exclusion, are listed in Appendix N. Articles were most frequently excluded because the study was not a controlled trial or a case series of 100 patients or greater, the study did not present original data, or the study did not report outcomes relevant to the study questions. The next most frequent reason for exclusion was that the article addressed only intracapsular cataract extraction (ICCE). Figure 2
provides a schematic representation of the results of the literature search, abstract review, and data abstraction process.Thirty-three percent of the articles included in this literature review were randomized controlled trials (Table 1). The remaining studies used non-randomized designs: 35 percent of studies were cohort studies; 21 percent were clinical series; five percent were case-control studies; and six percent were non-randomized controlled trials.
Forty-four studies reported on long-term control of IOP after treatment of patients with coexisting cataract and glaucoma, 15 studies reported on short-term control of IOP after treatment of patients with coexisting cataract and glaucoma, 22 studies reported on cataract development after surgical treatment of glaucoma, 16 studies reported on the efficacy of anti-fibrosis agents in surgical treatment of patients with coexisting cataract and glaucoma, seven studies compared outcomes of single- versus two-site surgery for patients with cataract and glaucoma, 11 studies compared outcomes of nuclear expression versus phacoemulsification for patients with cataract and glaucoma, four studies on compared outcomes of staged versus combined procedures for coexisting cataract and glaucoma, and three studies compared outcomes of trabeculectomy versus endoscopic laser treatment for patients with cataract and glaucoma.
Thirty-six of the 110 studies were randomized controlled trials, including six that presented data on long-term control of IOP after treatment of patients with cataract and glaucoma, five that presented data on short-term control of IOP after treatment of patients with cataract and glaucoma, 11 that presented data on cataract development after surgical treatment of glaucoma, 10 that presented data on the efficacy of anti-fibrosis agents in surgical treatment of patients with cataract and glaucoma, four that presented data on outcomes of single- versus two-site surgery for patients with cataract and glaucoma, two that presented data on outcomes of trabeculectomy versus endoscopic laser treatment for patients with cataract and glaucoma, none presenting data on outcomes of nuclear expression versus phacoemulsification for patients with cataract and glaucoma, and none that presented data on outcomes of staged versus combined procedures. A number of the studies addressed more than one of the study questions. A summary of key features of the studies reporting on each study question (number of eyes enrolled, mean patient age, proportion of patients African-American, mean length of study followup, and overall study quality) may be found in Appendix P.
A total of 131 articles had relevant information and appeared to be eligible for review. Twenty-one of these articles were excluded during the quality assessment, leaving 110 reports that underwent content assessment. The study designs were classified into five groups: 36 randomized controlled trials (33 percent of the 110), 38 cohort studies (35 percent), 23 clinical series (21 percent), six case-control studies (five percent), and seven non-randomized controlled trials (six percent).
Study quality category scores were calculated only for controlled trials (randomized and non-randomized) and for cohort studies, comprising 81 articles (74 percent of the 110). The overall quality data divided into study quality category are presented in Appendix Q. The overall mean quality score was 53 percent, with a range from three to 88 percent. The overall quality score for controlled trials (randomized or non-randomized) was higher than that for cohort studies, with the mean for controlled trials being 59 percent and for cohort studies 46 percent. Scores were higher for the controlled trials in all five categories.
Among the individual category scores for the 81 articles representing controlled trials and cohort studies, the mean for representativeness of the population was 58 percent, with a range from zero to 94 percent. Half the studies scored above 63 percent in this domain. For bias, the mean score was 28 percent, the lowest of the domains. Fully 25 of 81 studies (31 percent) had a score of zero, indicating no full randomization, no masking of observers measuring outcomes, and potential for groups to have major differences in key attributes at baseline (or no mention of these important methodological issues in the report). On the other hand, the description of the details of therapy was rated much higher, with a mean score of 65 percent. The statistics domain category had a mean score of 61 percent. The quality score for the outcomes category was intermediate at 53 percent.
Cataract Extraction alone versus Cataract Extraction and Glaucoma Procedure Randomized controlled trials. There were three randomized trials of good study quality (study quality scores of 67 percent8, 63 percent9, 73 percent10), comparing long-term IOP control in individuals with glaucoma undergoing phacoemulsification cataract extraction to those undergoing phacoemulsification cataract extraction combined with other glaucoma procedures. Anders and colleagues compared a no-stitch technique employing a radial incision in the base of the scleral tunnel to a scleral phacoemulsification cataract extraction approach.8 At 1 year, the straight phacoemulsification cataract extraction group had a reduction in IOP of 3.7 mm Hg with an average decrease of 0.5 medications. The combined group had an even larger decrease in IOP (7.70 mm Hg, p < .001) and required 1.5 fewer medications on average. Twenty-eight of 42 individuals in the phacoemulsification cataract extraction group still required medication at 1 year versus eight of 43 in the combined group. Storr-Paulsen, Pedersen, and Laugesen compared phacoemulsification cataract extraction to phacoemulsification cataract extraction plus trabeculectomy, but only randomized 10 eyes to each group.10 The median IOP decreased from 23 mm Hg to 16.5 mm Hg in the phacoemulsification cataract extraction group (range 12 to 18 mm Hg) as compared to a decrease from 20 mm Hg to 13 mm Hg (range 12 to 21 mm Hg) in the phacoemulsification cataract extraction plus trabeculectomy group. The distribution of IOP was not reported. A study by Gimbel and others compared phacoemulsification to phacoemulsification plus trabeculotomy.9 At 24 months, the mean decrease in the phacoemulsification alone group was 3.8 mm Hg versus 6.1 mm Hg (p = 0.001) in the combined group, but the phacoemulsification alone group was using fewer ocular hypotensive medications (mean of 0.4 versus 1.4). The results were reported on 51 patients in each group even though 53 were randomized per group.
Bobrow (quality score 82 percent) reported on patients who were randomized to extracapsular cataract extraction (ECCE) alone in one eye and ECCE plus trabeculectomy in the fellow eye (followed for an average of 87 months).11 Thirty-five patients were randomized to each group. Eyes randomized to combined procedures had a reduction in IOP of 8.2 mm Hg versus 4.3 mm Hg in the ECCE alone group (p < 0.001). Individuals in the combined group were taking on average 0.17 medications versus 0.77 in the ECCE alone arm. No patient required additional surgical procedures to control eye pressure.
Non-randomized studies. One study (quality score 42 percent) retrospectively reviewed the charts of 21 glaucoma patients undergoing combined phacoemulsification cataract surgery and trabeculectomy and compared the IOP lowering effect in 35 eyes of glaucoma patients who underwent phacoemulsification alone.46 IOP decreased a mean of 3.4 mm Hg in the phacoemulsification cataract extraction alone group as compared to 6.8 mm Hg in the combined group at 6 months with no difference in the number of ocular hypotensive medications being used.
Conclusion: Studies comparing cataract extraction alone to cataract extraction and a glaucoma procedure found that while IOP is lowered by cataract extraction alone by about 4 mm Hg, the addition of a glaucoma procedure lowers IOP even more and further may reduce the need for medical therapy. [Evidence Grade: A]
Two additional randomized trials quality scores 46 percent12 and 45 percent13 looked at aspects of the combined procedure that may influence long-term IOP. Kim, Jung, and Hwang randomized individuals to phacoemulsification cataract extraction plus either a no-stitch trabeculectomy (11 patients), a T-flap trabeculectomy (eight patients) or a trabeculectomy with sutures (nine patients).12 All three groups had a similar 6 month IOP (around 13 mm Hg), but the group with sutures had a much higher baseline eye pressure. Conclusions are difficult to draw from this study because of small sample size and large baseline differences between the study groups. Hugkulstone and colleagues reported on seven eyes undergoing ECCE plus trabeculectomy with a standard intraocular lens (IOL) and nine patients undergoing the same procedure with a heparin-coated IOL at 24 months.13 The mean drop in IOP was 3.8 mm Hg in the standard IOL group and 9.7 mm Hg in the heparin-coated IOL group (p < 0.05).
Non-randomized studies. We found no non-randomized studies addressing this topic.
Randomized controlled trials. We found no randomized controlled trials addressing this topic.
Non-randomized studies. Four articles of varying quality compared limbus- to fornix-based incisions (study quality scores of 66 percent14, 48 percent15, 41 percent16, and 68 percent17). Each of the four articles comparing limbus- to fornix-based incisions for combined procedures reported no difference in long-term IOP control using the two approaches. A study by Murchison and Shields was the only study reporting on ECCE plus trabeculectomy.17 These authors reported no difference in long-term IOP control comparing 22 limbus-based cases to 25 fornix-based cases followed for a mean of 18 months. The remaining three studies looked at phacoemulsification cataract extraction plus trabeculectomy. The first study, by Shingleton and colleagues reported on patients undergoing bilateral phacoemulsification cataract extraction plus trabeculectomy.14 First eyes (44) were assigned to limbus-based conjunctival flaps and fellow eyes (44) to fornix-based flaps. The results were virtually identical in the two groups, with a mean decrease of 6.1 mm Hg. Studies by Berestka and Brown15 and Tezel and colleagues16 reported on retrospective chart reviews and found similar results in 215 and 45 cases respectively. Both studies reported a mean decrease in IOP of approximately 8 mm Hg.
While the small sample size of the studies makes it unlikely that real differences could be deleted, the very similar IOP results indicate that large differences in effectiveness are unlikely to be present.
Conclusion: Limbus- and fornix-based approaches to combined procedures appear to have similar IOP-lowering effects. [Evidence Grade: B]
Randomized controlled trials. We found no randomized controlled trials addressing this topic.
Non-randomized studies. Four non-randomized controlled studies of varying quality (study quality scores of 65 percent18, 35 percent19, 58 percent20, and 60 percent21) compared larger versus smaller incisions during combined phacoemulsification cataract extraction plus trabeculectomy procedures. Lyle and Jin reported a slightly greater decrease in IOP with a 3 mm incision (104 eyes) versus a 6 mm incision (112 eyes) (8.9 mm Hg versus 6.8 mm Hg) but the difference was not statistically significant.18 Wand reported on a study using mitomycin-C and found no significant difference between the two groups (mean decrease in IOP of 7.3 mm Hg in the smaller incision group, (33 eyes) versus 9.1 mm Hg in the larger incision group, (24 eyes).20 The third article, by Kosmin, Wishart, and Ridges found no difference in outcomes comparing small incision surgery using silicone lenses (33 eyes) to larger incision surgery (33 eyes) using poly(methyl methacrylate) (PMMA) lenses (mean IOP decreased 7.7 mm Hg in the small incision group versus 6.7 mm Hg in the larger incision group).21 Vyas and colleagues reported that IOP decreased by 7.6 mm Hg in the small incision group (25 eyes) and 7.8 mm Hg in the larger incision group (25 eyes) at 13 months.19 While the majority were on at least one medication preoperatively, 49 of 50 were taking no medications postoperatively.
The small sample size of many of these studies means that real differences may have been missed. However, the results across studies consistently show no difference.
Conclusion: Phacoemulsification cataract extraction plus trabeculectomy resulted in a decrease in IOP of about 8 mm Hg with both 3 mm and 6 mm wound sizes. [Evidence Grade: B]
Randomized controlled trials. We found no randomized controlled trials addressing this topic.
Non-randomized studies. Three non-randomized studies (quality scores of 21 percent22, 83 percent23, and 63 percent24) compared IOP results using phacoemulsification cataract extraction plus trabeculectomy to trabeculectomy alone. Bellucci and colleagues compared 100 trabeculectomies to 200 phacoemulsification cataract extraction plus trabeculectomies and found that trabeculectomy alone resulted in a larger decrease in IOP than the combined procedure (11.2 mm Hg versus 3.1 mm Hg p < .01).22 However, initial IOP was much higher in the trabeculectomy alone group (26.4 mm Hg versus 21.8 mm Hg). A cohort study by Caprioli and others addressed this question, comparing 40 cases of phacoemulsification cataract extraction plus trabeculectomy (two-site) to trabeculectomy alone.23 These authors reported that IOP decreased more in the trabeculectomy alone group (10.3 mm Hg versus 7.4 mm Hg), and that a higher proportion achieved its target pressure in the trabeculectomy alone group (88 percent versus 72 percent). Derick, Evans, and Baker reported a slightly larger decrease in IOP in the trabeculectomy alone group (10.6 mm Hg versus 8.9 mm Hg), with similar medication requirements (42 eyes in each group).24
Conclusion: Non-randomized studies indicate that phacoemulsification cataract and trabeculectomy combined results in slightly worse IOP control than trabeculectomy alone. [Evidence Grade: C]
Randomized controlled trials. We identified no randomized controlled trials addressing this question.
Non-randomized studies. Seven studies reported on the results of phacoemulsification cataract extraction plus trabeculectomy.25–31 Four studies were case series and so quality scores were not calculated.25–28 The three remaining studies were selected for review because they compared two different approaches to combined phacoemulsification cataract surgery. These studies provide case series data relevant to the current question, but have fewer than 100 subjects enrolled and therefore may not be as informative.
With regard to the four large case series, Crandall did not report long-term IOP results, but found that four of 100 eyes (98 Caucasian) required subsequent IOP-lowering surgery.25 Jayamanne, Kostakis, and Phelan followed 100 eyes for a mean of 11 months and found that 96 percent had an IOP less than 21 mm Hg with a mean decrease in IOP of 7.6 mm Hg.26 The third large case series, by Yang and colleagues reported on 182 eyes (12 percent of subjects were African-American), all of whom received mitomycin-C with a mean followup of 17 months.27 The mean IOP decrease was 7.4 mm Hg and 95 percent achieved an IOP less than or equal to 21 mm Hg. The last case series employed a no-stitch trabeculectomy combined with a trabeculectomy along with phacoemulsification cataract extraction and reported a mean decrease in IOP of 5.1 mm Hg at 2 years.28
Arnold compared a no-stitch group (20 eyes) to a two-stitch group (20 eyes) with similar 24 month decreases in IOP (-7.5 mm Hg and -8.0 mm Hg, respectively).29 Patients required virtually no medications, and all patients had an IOP less than or equal to 21 mm Hg. Crestani and others reported on the effect of phacoemulsification cataract extraction plus trabeculectomy using either a punch or a standard trabeculectomy and found no difference in IOP-lowering at unknown followup (-12.6 mm Hg with the punch versus -14.2 mm Hg with the conventional approach).30 The number of eyes was not reported. Nielsen compared two techniques of performing phacoemulsification cataract extraction plus trabeculectomy, with one approach reducing IOP by 14 mm Hg (10 eyes) and the other by 8 mm Hg (26 eyes).31
Conclusion: Phacoemulsification cataract extraction plus trabeculectomy in these series as well as in the comparative studies discussed above, resulted in an IOP decrease of about 8 mm Hg in individuals followed for a mean of 1 to 2 years. [Evidence Grade: C]
Randomized controlled trials. We identified no randomized controlled trials addressing this question.
Non-randomized studies. Two cohort studies (quality scores, 49 percent32 and 66 percent33) compared ECCE plus trabeculectomy to trabeculectomy alone. Naveh and colleagues reported that IOP decreased by 9.8 mm Hg in the combined group (40 eyes) compared to 12.1 mm Hg in the trabeculectomy alone group (38 eyes) at 18 months, with an even greater disparity at the last reported followup (mean of -8.0 mm Hg versus B 14.7 mm Hg).32 The combined group also required an average of 2.1 medications versus 0.5 in the trabeculectomy alone group. Yu and others also reported greater IOP lowering with trabeculectomy alone (mean of 11.4 mm Hg in 24 eyes) than with a combined procedure (mean of 7.5 mm Hg in 21 eyes) with similar medication use in the two groups.33
Conclusion: Trabeculectomy alone appears to lower IOP more than combined ECCE and trabeculectomy. [Evidence Grade: B]
Randomized controlled trials. We identified no randomized controlled trials addressing this question.
Non-randomized studies. In addition to the two cohort studies sited above, two studies reported on the results of ECCE plus trabeculectomy. One study was a clinical series and quality scores were not calculated for this study.34 The one remaining study had a quality score of 36 percent.35 In a case series, McCartney and colleagues reported on a mix of 102 open-angle glaucoma (OAG) and angle closure glaucoma (ACG) subjects followed for a mean of 17 months.34 Intraoperative pressure decreased an average of 6.2 mm Hg and medication requirements declined as well. The other article compared ECCE plus trabeculectomy with and without intraocular lens (IOL) implantation with a mean IOP decline of 6.3 mm Hg in the group with an IOL (23 eyes) followed for an average of 11 months.
Conclusion: ECCE and trabeculectomy combined results in a 6 to 8 mm Hg decline in subjects followed for 1 to 2 years. [Evidence Grade: C]
Non-randomized studies. Seven articles reported on the IOP-lowering effect of ECCE alone in glaucoma patients.36–42 Quality scores were calculated for only two articles (40 percent).37 The first article by McGuigan and colleagues reported that all 40 subjects undergoing ECCE had IOP less than 24 mm Hg prior to surgery.37 At an average followup of 18 months, 19 eyes were using the same number of medications, 11 required additional medications and 10 required fewer medications. Calissendorff and Hamberg- Nystrom reported the results of a case series with 74 remaining subjects followed for 3 years.38 The mean decrease in IOP among 11 ocular hypertensives (who had higher baseline IOP) was 5.5 mm Hg, while among 63 glaucoma patients it was 2.1 mm Hg. In the third case series, Cinotti and others reported on 160 glaucoma patients undergoing ECCE with a mean followup of 1 year finding a decrease in IOP of 1.4 mm Hg.39 Medication requirements did not change. Onali and Raitta reported on 37 patients with OAG and 66 with pseudoexfoliation glaucoma with a mean followup of over 2 years.40 IOP declines were similar in the two groups with a mean decrease of 1.5 mm Hg. Kooner and others reported on 91 ECCE cases in glaucoma patients without prior trabeculectomy.41 With a mean followup of 18 months, IOP decreased an average of 2.2 mm Hg; 81 percent were on the same number of medications, 13 percent were on more medications, and eight percent were on fewer medications at 1 year. Gunning and Greve looked at 102 with ACG or OAG requiring cataract extraction, including previous filters.42 While the overall IOP lowering was 5.2 mm Hg, the ACG group declined 7.3 mm Hg while the OAG group dropped 2.6 mm Hg. The final case series reported on twelve eyes undergoing ECCE with IOP implantation with a mean decrease in IOP of 1 mm Hg with over 1 year of followup.36
Conclusion: ECCE alone in patients with glaucoma resulted in a modest reduction in IOP (about 1 to 3 mm Hg) at 1–2 years followup, with little change in the use of IOP-lowering medications. [Evidence Grade: C]
Non-randomized studies. In addition to the four controlled trials that included a study group of glaucoma patients undergoing phacoemulsification cataract extraction (PECE) alone, three articles reported the IOP-lowering effect of phacoemulsification cataract extraction alone in patients with glaucoma.43, 44, 45 Study quality scores were calculated for two of the three articles (50 percent43 and 60 percent44). Hayashi and colleagues reporting on a study from Japan, reported results separately for individuals with ACG (77 eyes) and OAG (73 eyes).43 At 12 months, the average IOP decreased by 6.4 mm Hg in the ACG group and 4.1 mm Hg in the OAG group. Shingleton and others reported on glaucoma suspects (75 eyes) and individuals with glaucoma (71 eyes).44 At 1 year IOP decreased a mean of 1.9 mm Hg in the glaucoma suspects and 1.1 mm Hg in those with glaucoma, with glaucoma patients requiring fewer medications. Perasalo reported on 127 eyes followed up at 1 year and found a mean decrease in IOP of 1.8 mm Hg.45
Conclusion: Phacoemulsification cataract extraction alone in glaucoma patients documented a 1 to 4 mm Hg decrease in IOP at 1 to 2 years of followup. [Evidence Grade: C]
Non-randomized studies. Five articles looked at the effect of cataract extraction on IOP in patients with preexisting filtering blebs.23, 46, 42, 47 Quality scores are available for two of the four articles (83 percent23 and 37 percent46). The first article, by Caprioli and colleagues, reported results of a survival analysis in a matched case-control study comparing individuals with existing blebs (40 eyes) to those with existing blebs undergoing clear cornea phacoemulsification cataract extraction (40 eyes) and found no difference in IOP control over time.23 Chen retrospectively reviewed 115 patients with functioning filtering blebs in whom either ECCE (58) or PECE (57) was performed using both clear corneal and scleral approaches.48 For the group as a whole, 30.4 percent required either additional medication or bleb needling to maintain IOP control at a mean followup of 21 months. An additional eleven eyes (9.6 percent) required additional glaucoma surgery. A Cox proportional hazards model found the risk of requiring glaucoma surgery was 3.0 (95 percent confidence interval: 0.7, 12.8) for those undergoing ECCE as compared to PECE. No control group was enrolled to assess the baseline rate of bleb failure in this population.
Manoj and colleagues compared ECCE to PECE in patients with preexisting filtering blebs.46 IOP increased significantly (1.9 mm Hg) in the ECCE group (34 eyes), but remained unchanged in the phacoemulsification cataract extraction group (21 eyes). Medication requirements were also greater in the ECCE group. Gunning and Greve reported a mean decrease of 1.5 mm Hg (not significant) in individuals undergoing ECCE with preexisting filters.42 Finally, Burratto and Ferrari studied 102 individuals undergoing either ECCE or phacoemulsification with preexisting blebs.47 Although the data presented were limited, it appeared that 8 months after cataract surgery the IOP was similar to baseline.
Conclusion: All but one study indicate that cataract extraction (both ECCE and PECE) has a minimal impact on the functioning of preexisting filtering blebs. The data are insufficient to determine if cataract extraction alters the function of preexisting filtering blebs. [Evidence Grade: I]
Other studies. Other authors reported on unique approaches to the glaucoma patient with cataract. Montgomery and Gills assessed the IOP-lowering effect of ECCE plus cyclodialysis (238 eyes), reporting that, at a mean of 15 months, 86 percent had Aacceptable@ IOP with 54 percent on no medications.49 Honjo and others (study quality score 13 percent) reported the results of ECCE (25 eyes) or phacoemulsification cataract extraction plus trabeculotomy (22 eyes).50 IOP was lowered in both groups (mean of 9.4 mm Hg in the ECCE group and about 9 mm Hg in the phacoemulsification cataract extraction group) with a mean followup of 29 months. The PECE group used two different incisions for the trabeculotomy, but results were similar in the two groups.
Randomized controlled trials. There were five randomized trials that reported on short-term IOP control in individuals with glaucoma undergoing cataract surgery. Quality scores for the studies were good (54 percent51, 73 percent10, 63 percent9, 73 percent52 and 66 percent53). Three studies compared cataract extraction to cataract extraction plus a glaucoma procedure.51, 10, 9 A study by Krupin, Feitl, and Bishop used an unusual randomization scheme.51 Individuals with controlled glaucoma were randomized to either ECCE or ECCE plus trabeculectomy. Those with uncontrolled glaucoma were assigned to ECCE plus trabeculectomy or ECCE plus trabeculotomy. Ultimately, 26 eyes were randomized to ECCE, 33 to ECCE plus trabeculectomy, and nine eyes to ECCE plus trabeculotomy. The analysis combined the ECCE plus glaucoma surgery groups even though the trabeculotomy results were not as good as those with trabeculectomy. ECCE alone resulted in a 15.3 mm Hg average increase in IOP with 22 of 26 eyes showing an increase in IOP on postoperative day 1. The ECCE plus glaucoma surgery groups had a mean decrease of 6.1 mm Hg in IOP with 12 of 42 showing an increase in IOP on postoperative day 1.
Storr-Paulsen, Pedersen, and Laugesen compared PECE alone to PECE plus trabeculectomy, but only randomized 10 eyes to each group.10 Timolol was used at the end of surgery and acetazolamide was given 6 hours after surgery. The median IOP decreased from 23 mm Hg to 20.5 mm Hg in the phacoemulsification cataract extraction group (range 0 to 32 mm Hg) as compared to a decrease from 20 mm Hg to 8 mm Hg (range 0 to 35 mm Hg) in the phacoemulsification cataract extraction plus trabeculectomy group.
The third study, by Gimbel and colleagues, compared phacoemulsification to phacoemulsification plus trabeculotomy.9 The authors did not report using any ocular hypotensive at the end of the case. The results at 1 day were similar between the two groups with a mean decrease in IOP of 2.0 mm Hg in the phacoemulsification cataract extraction arm and 2.2 mm Hg in the phacoemulsification cataract extraction plus trabeculotomy arm. The results were reported on 51 patients in each group even though 53 were randomized per group.
Conclusion: There are insufficient data to determine if adding a glaucoma procedure to cataract surgery results in lower IOP than preoperatively or lower IOP than cataract surgery alone 24 hours after surgery. [Evidence Grade: I]
The fourth and fifth randomized trials reporting on short-term IOP control assessed medical approaches to prevent IOP spikes in glaucoma patients undergoing ECCE plus trabeculectomy.52, 53 A study reported by Robin demonstrated that apraclonidine given at the end of surgery resulted in lower IOP at 1 day after surgery in individuals undergoing ECCE plus trabeculectomy.52 The group receiving apraclonidine (59 eyes) had a mean decrease in IOP of 12.4 mm Hg versus 0.6 mm Hg in the control group (50 eyes) (all patients were given pilocarpine 2 percent every 6 hours after surgery, and those on beta-blockers prior to surgery continued taking them).
The fifth study, by West and others compared three treatment arms: acetazolamide alone 1 hour prior to the beginning of the case (10 eyes), miochol intraoperatively (11 eyes), and preoperative acetazolamide and intraoperative miochol (15 eyes).53 Acetazolamide alone and miochol alone both were unable to prevent IOP elevation (mean increase of 8.9 mm Hg and 6.3 mm Hg, respectively), while the combination led to an average decrease in IOP of 0.7 mm Hg on the first day after surgery.
Conclusion: Individual studies indicate that apraclonidine or a combination of miochol and acetazolamide blunt the IOP rise in glaucoma subjects undergoing combined ECCE and trabeculectomy who were already on maximal medical management. [Evidence Grade: C]
Non-randomized studies. The case series and non-randomized studies in this table reported on a wide range of surgical techniques. Six studies of varying quality (quality scores of 24 percent54, 60 percent21, 41 percent16, 65 percent18, 63 percent24 and 60 percent55) reported the results of PECE on short-term IOP control among glaucoma patients. Bellucci and Luraschi compared 10 patients in each group receiving either a one- or two-site combined PECE plus trabeculectomy.54 The IOP decrease was greater in the two-site group (mean of 20.6 mm Hg versus 10.5 mm Hg). Kosmin, Wishart and Ridges compared results using silicone versus PMMA lenses in PECE plus trabeculectomy patients and found a greater decrease in IOP at 1 day using the PMMA lenses (8.7 mm Hg versus 3.4 mm Hg). 21 However, Healon GV was used in silicone cases as opposed to Healon in PMMA cases. Lyle and Jin compared 3 mm (104 eyes) versus 6 mm (112 eyes) incisions in PECE plus trabeculectomy cases and found similar declines in IOP on postoperative day 1 (mean of 7.8 mm Hg versus 9.9 mm Hg).18 Tezel and others compared limbus (151 eyes) versus fornix-based (64 eyes) flaps in PECE plus trabeculectomy and found a similar immediate decrease in IOP (mean of 7.4 mm Hg versus 7.3 mm Hg).16 The fifth article by Derick, Evans, and Baker found a mean decrease of 6.2 mm Hg with PECE plus trabeculectomy on postoperative day 1.24 Only five out of 42 individuals had an IOP greater than 25 mm Hg. Finally, Shingleton and colleagues, looked at PECE in glaucoma suspects (75 eyes) and glaucoma patients (71 eyes).44 The IOP decreased a mean of 1.0 mm Hg in the suspects and increased a mean of 1.0 mm Hg in the glaucoma patients (all patients received levobunolol at the end of the case).
Conclusion: These articles on PECE in glaucoma patients indicate that combined procedures result in IOP decreases at 1 day postoperatively. [Evidence Grade: C]
Conclusion: Data are insufficient to determine the impact of phacoemulsification cataract extraction alone on short-term IOP control. [Evidence Grade: I]
Four additional articles (with study of quality scores 36 percent56, 40 percent37, 68 percent17 and 30 percent57) addressed the short-term impact of ECCE on IOP in glaucoma subjects. Vu and Shields found similar IOP rises with ECCE (25 eyes) as with intracapsular cataract extraction (ICCE) (25 eyes).56 McGuigan and colleagues reported a mean IOP rise of 10.2 mm Hg for glaucoma patients undergoing ECCE (80 eyes) whereas nine of 10 patients undergoing a combined procedure had an IOP less than 10 mm Hg on the first postoperative day.37 In a study comparing limbus- (22 eyes) and fornix-based flaps (25 eyes), Murchison and Shields found that a slightly higher proportion undergoing the limbus-based procedure had a decrease in IOP (100 percent versus 76 percent).17 McKinzie and Boggs compared intracameral acetylcholine (8 eyes) to intracameral carbachol (8 eyes) at the end of the case and found that IOP increased by 16.4 mm Hg in the acetylcholine group while it decreased by 0.1 mm Hg in the carbachol group.57
Conclusion: Combined ECCE and trabeculectomy results in lower IOP on the first postoperative day. [Evidence Grade: C]
Conclusion: ECCE alone in glaucoma patients appears to increase IOP on the first postoperative day. [Evidence Grade: C]
Randomized controlled trials. Two recent randomized controlled trials suggest that glaucoma surgery is associated with higher rates of development and progression of cataract. Persons randomized to have IOP lowered by 30 percent from baseline in the Collaborative Normal Tension Glaucoma Study (study quality score of 67) many of whom received incisional glaucoma surgery, had an incidence of cataract (23 of 61, or 38 percent) which was significantly greater than for the group not randomized to undergo pressure reduction (11 of 79, or 14 percent, p = 0.0075).58, 59 Likewise, in the Advanced Glaucoma Intervention Study (study quality score of 70 percent), the rate of cataract operations among black subjects randomized to trabeculectomy first (TAT) (34.6 percent) was significantly (p < 0.001) higher than among black subjects randomized to receive argon laser trabeculoplasty first (ATT) (22.5 percent).60 Although the cumulative probability of cataract surgery for white patients randomized to TAT (37.1 percent) was also higher than for ATT (29.7 percent), the difference was not statistically significant (p = 0.097).
Non-randomized studies. Quality scores were not calculated for eight of the non-randomized studies reporting data relevant to this question.61–68 The studies for which quality scores were determined were of varying quality (43 percent69, 75 percent70, 73 percent71, 85 percent72, 35 percent73, 75 percent74 and 22 percent75). In a non-randomized trial in India, in which a subject's eye with higher IOP received trabeculectomy and the other eye medications, Dastur reported that 12 of 70 eyes required cataract extraction within 3 years after trabeculectomy, whereas no cataract surgery was performed in the medically treated eyes.69
In a prospective study of visual acuity loss after trabeculectomy involving 508 eyes of 440 patients, Costa reported lens opacification as the leading cause of two or more lines of vision loss in this group.61 Cataract progression/formation was also the most common postoperative complication (20 percent), and the leading cause of postoperative vision loss in Cheung's series of 157 trabeculectomies with mitomycin-C.62 In his clinical series of trabeculectomies, Watson found presence of cataract prior to glaucoma surgery to be the major determinant of increase in cataract.63 Cataract was not related to preoperative IOP, reduction in IOP, operative complications other than flat chamber with corneal touch, the type or frequency of preoperative medicines, or the type of glaucoma. Tornqvist64 found a doubling of the visual angle due to cataract in 37 percent of persons after trabeculectomy after 5 years, while Jerndal65 found that 10 percent of subjects required cataract extraction after trabeculectomy over a similar period of followup.
Lamping and colleagues reviewed 181 eyes that underwent full-thickness glaucoma and 71 eyes undergoing trabeculectomy that had at least 2 years follow-up.69 they found that cataract extraction was performed in 34 percent of full thickness procedures and in 21 percent of the trabeculectomies, a difference that was of borderline statistical significance (p = 0.06). Although the amount of cataract surgery seems to exceed that which would have occurred in eyes no undergoing glaucoma surgery, the lack of a control group not undergoing surgery prohibits drawing firm conclusions.
Several of the reviewed studies considered other factors associated with glaucoma surgery which might increase the risk of lens opacity. In his randomized trial of mitomycin-C concentration and duration in India, Robin reported no eyes with a 2-line decrease in vision or requiring cataract surgery among subjects randomized to receive trabeculectomy with placebo, while the age-adjusted relative risk for progression of cataract was highest in the two groups receiving the highest doses of mitomycin-C: 0.4 mg/ml for 2 minutes (relative risk 2.17, p = 0.09) and 0.2 mg/ml for 4 minutes (relative risk 3.18, p = 0.01).70 Kitazawa found cataract progression (18 percent) exclusively in subjects randomly allocated to receive mitomycin-C (0.2 mg), with no lens change reported in those receiving a 10-fold lower dose.71 This small trial included a total of 19 eyes. Rates of cataract progression did not differ significantly between subjects randomized to receive a mitomycin-C concentration of 0.4 mg/cc (three of four) and those receiving 0.2 mg/cc (four of five), in a small trial reported by Sanders and colleagues.72 Similarly, no difference in rate of cataract progression was reported between patients randomized to daunorubicin (three of 13) and those treated with 5-fluorouracil (three of 14) in Demailly's randomized trial.73
Asamoto reported significantly higher rates of anterior subcapsular cataract among subjects undergoing trabeculectomy in which the anterior chamber was reformed with air, as opposed to balanced salt solution (BSS), sodium hyaluronate, or spontaneous reformation.66 There were 11 permanent cataracts in eyes reformed with air, but only two were visually significant, and only one required surgery.
Other surgical treatments for glaucoma are also associated with postoperative cataract although controlled data do not generally exist to demonstrate whether this risk is increased compared to non-operated subjects. Progression of cataract was reported in seven percent of subjects after deep sclerostomy with collagen implant after a mean 18 month followup by Karlen.67 The rate of cataract progression varied between two to four percent after adult trabeculotomy in two studies reported by Tanihara.68 Three eyes receiving argon laser trabeculoplasty underwent cataract extraction, as compared to one (3 percent) in a small, controlled, retrospective study by Tuulonen.76 Higher rates of cataract were reported in association with full thickness glaucoma surgeries as opposed to trabeculectomy by Wilson77, Lewis78, and Blondeau.74 Spaeth found roughly equivalent rates of cataract progression between the two procedures in a smaller (n=15) study.75
Glaucoma surgery appears to be associated with increased risk of postoperative visually significant cataract, which may require cataract surgery. The risk of postoperative cataract may be increased with age, use of mitomycin-C, intraoperative use of air to reform the anterior chamber, postoperative flat chamber, and full thickness glaucoma surgery.
Conclusion: Glaucoma surgery is associated with increased risk of postoperative cataract. [Evidence Grade: A]
Randomized controlled trials. The overall quality of these studies (mean score 59 percent, range from 48 percent to 67 percent) exceeded those of the average study in this report (mean score 53 percent).
Three groups have conducted randomized controlled trials evaluating the effect of mitomycin-C in combined cataract and glaucoma surgery.79–83 Cohen and colleagues, performed a study (72 eyes total) using a high dose of mitomycin-C (0.5 mg/ml for 2.5 minutes) versus placebo at the time of phacoemulsification and trabeculectomy.79 They found that at 12 months, with data for 52 eyes, the mitomycin-C group had greater IOP lowering (mean of 7.7 mm Hg versus 3.1 mm Hg, p=0.001), was on fewer medications (mean of 0.6 versus 1.4 per eye, p=0.012), and had larger filtration blebs. Carlson and others, performed a smaller study (29 eyes total), also using a high dose of mitomycin-C (0.5 mg/ml for 3.5 minutes) versus placebo at the time of phacoemulsification and trabeculectomy.80 At one year, the IOPs were lower (mean of 5.6 versus 2.6, p=0.11) and fewer eyes needed medications (zero of 14 eyes versus five of 15 eyes, p=0.04) in the mitomycin-C group than in placebo group. Throughout the course of the study, eyes treated with mitomycin-C had an IOP that averaged 3 mm Hg less than the placebo-treated eyes (p=0.04).
Shin's group conducted a number of studies on this topic.81–83 Unlike the other studies, their trials included a significant proportion of African-Americans. There may be some overlap in the eyes reported in their papers. The authors distinguished between eyes without previous glaucoma surgery, and those with previous glaucoma surgery.
In the largest study by Shin and colleagues, 96 eyes without previous glaucoma surgery received mitomycin-C 0.5 mg/ml for varying lengths of time at the time of combined phacoemulsification and trabeculectomy and 101 eyes without previous glaucoma surgery received placebo.82 Using life table analysis, no difference (p = 0.12) was found in the success rate of glaucoma surgery between the eyes receiving mitomycin-C (approximately 65 percent at 2 years) and those receiving placebo (approximately 50 percent at 2 years). However, there was a trend for greater success (65 percent) at 30 months in the mitomycin-C treated group than in the control group (50 percent). Using life table analysis mitomycin-C improved the success rate in subgroups including eyes of African-Americans (p = 0.007), in eyes on two or more glaucoma medications (p = 0.018) and in eyes with IOP of 20 mm Hg or greater (p = 0.006).
In a second randomized trial from Shin's group, Lemon and colleagues randomized 69 eyes without previous glaucoma surgery receiving 0.5 mg/ml of mitomycin-C for only 1 minute at the time of combined phacoemulsification and trabeculectomy to either a fornix- or limbus-based conjunctival flap, and found no difference in IOP lowering (mean 3.3 mm Hg) or complications between the two groups.83
In a third study, Shin, Kim, Shoth and colleagues, randomized eyes with primary open-angle glaucoma (POAG) and failed previous glaucoma filtration surgery to phacoemulsification and trabeculectomy with or without mitomycin-C (0.5 mg/ml for one minute).81 Using life table analysis, the use of mitomycin-C appeared to increase the success rate of surgery (p = 0.02). At 14 months, for example, the success rate was approximately 60 percent in the mitomycin-C treated group and 25 percent in the control group.
Non-randomized studies. Budenz and colleagues in a study comparing 5-fluorouracil, mitomycin-C, and no anti-fibrosis agent, failed to find a significantly greater degree of IOP lowering in eyes that received mitomycin-C at the time of phacoemulsification and trabeculectomy (mean 4.6 mm Hg) than in those receiving no anti-fibrosis agent (mean 3.4 mm Hg).84 Their results are difficult to interpret because the IOP fell less in the 5-fluorouracil group than in the group that did not receive an anti-fibrosis agent. The number of medications required after surgery was not significantly different between the mitomycin group (mean 0.4) and the no anti-fibrosis group (mean 0.7). However, overall the mitomycin-C group had lower IOP and required fewer medications.
Liu and colleagues in Taiwan compared 22 eyes that underwent combined phacoemulsification and trabeculectomy with 0.2 mg/ml mitomycin-C for 3 minutes to 18 eyes that did not receive mitomycin-C at the time of combined cataract and glaucoma surgery.85 The IOP at 1 year was lower in the mitomycin-C group (mean 10.9 mm Hg versus 14.6 mm Hg, p=0.004), but because the mitomycin-C group had a lower preoperative IOP (mean 16.1 mm Hg versus 18.7 mm Hg), the decrease in IOP was only slightly greater in the mitomycin-C group, (mean 5.16 mm Hg versus 4.14 mm Hg for the controls). The authors did not indicate whether this difference was statistically significant.
El-Sayyad and colleagues compared combined phacoemulsification and cataract extraction with mitomycin-C to trabeculectomy with mitomycin-C followed at a later date by phacoemulsification (mean delay 3.2 months).86 In both instances, the IOP fell by a mean of greater than 13 mm Hg.
Shin, Kim, Ren and colleagues evaluated the incidence of posterior capsular opacification in eyes without previous glaucoma surgery receiving mitomycin-C or placebo at the time of combined phacoemulsification and trabeculectomy.87 Using life table analysis they found that YAG capsulotomy was performed less frequently in the eyes that received mitomycin-C (p = 0.004). For instance, the proportion of eyes receiving capsulotomy by 4 years after surgery was approximately 60 percent in the placebo group and 40 percent of the mitomycin-C group.
In summary, the literature suggests that overall, the use of mitomycin-C in combined procedures results in lower IOP. Two fairly high quality randomized controlled trials found an IOP-lowering effect of mitomycin-C. A larger, high quality study (and the only one with a significant number of African-Americans) failed to find an overall effect, but did demonstrate an effect when important subgroups were analyzed.87 No serious side effects attributable to mitomycin-C have been seen, and it may reduce the rate of posterior capsular opacification.
Conclusion: Overall, the use of mitomycin-C in combined cataract and glaucoma surgery results in lower IOP. [Evidence Grade: B]
Randomized controlled trials. There were four randomized controlled trials reporting the effect of 5-fluorouracil in combined cataract and trabeculectomy surgery.88–91 The mean overall quality score of these studies (48 percent) was close to the score for all the randomized trials evaluated in this report (53 percent). In two studies, ECCE was used. In the first study, in a largely African-American population, Hennis and Stewart randomized 32 eyes undergoing combined ECCE and trabeculectomy to receive up to 14 daily injections of 5-fluorouracil or not.88 With short followup of 3 months, no difference could be detected in IOP (power of finding a 4 mm difference of 92 percent), number of glaucoma medications, or bleb appearance between the two groups. In a second study, Wong and colleagues randomized 24 predominantly Caucasian patients to receive five injections of 5 mg during the first 2 postoperative weeks and 20 patients to not receive 5-fluorouracil.89 IOP was lowered by a mean of 4 mm Hg in both groups, and there were no important differences between them. The authors did not comment on the study's power to detect a difference between the two groups.
Phacoemulsification combined with trabeculectomy was used in the other two randomized controlled trials. O'Grady and others, in a predominantly African-American population, randomized 74 patients to either receive or not receive 5-fluorouracil (mean of 5 injections of 5 mg during the first week after surgery) following single-site combined phacoemulsification and trabeculectomy.90 No beneficial effect of 5-fluorouracil was seen (p values at all time points >0.05, but no power calculations were reported). In an Italian population, Gandolfi and Vecchi randomized 24 eyes undergoing a limbus-based trabeculectomy and clear corneal phacoemulsification to receiving 5 once-weekly injections of 5 mg of 5-fluorouracil.91 At 1 year, 10 of 12 eyes in the 5-fluorouracil group had IOPs of 15 mm Hg or less compared to one of 12 in the control group (p=0.001, Fisher's exact test), suggesting that 5-fluorouracil injections were beneficial.
Non-randomized studies. Budenz and others compared the results of eyes undergoing trabeculectomy and phacoemulsification and receiving either mitomycin-C, 5-fluorouracil injections, or no anti-fibrosis agent.84 The 5-fluorouracil doses were not standardized. At 1 year, the mean IOP lowering was 2.7 mm Hg in the 5-fluorouracil group and 3.4 mm Hg in the control
group. The authors concluded that 5-fluorouracil was not beneficial in combined phacoemulsification and trabeculectomy.
Cohen and others also compared the results of eyes randomized to either receive or not receive 5-fluorouracil at the time of combined cataract and glaucoma surgery.92 Some eyes had undergone previous glaucoma surgery, and the techniques for cataract surgery varied. A mean of 17 mg 5-fluorouracil was administered during the first 2 postoperative weeks. The mean fall in IOP with about 6 months followup was 4.5 mm Hg in the control group and 4.1 mm Hg in the group receiving 5-fluorouracil.
Hurvitz compared 38 eyes receiving 5-fluorouracil and combined phacoemulsification and trabeculectomy with 55 eyes that underwent the same procedure, but without 5-fluorouracil.93 At 1 month the IOP was a mean of 1.9 mm Hg lower in the phacoemulsification/5-fluorouracil group than in the phacoemulsification alone group (p=0.08). At 1 year, the decrease in IOP was slightly greater (mean of 6.2 mm Hg) in the group receiving 5-fluorouracil than in the group not receiving 5-fluorouracil (mean of 5.3 mm Hg) (no statistical analysis provided).
Donoso and Rodriguez compared the results of single incision, combined trabeculectomy and phacoemulsification with that of a matched group of eyes undergoing phacoemulsification after trabeculectomy.94 In both groups, a single intraoperative application of 5-fluorouracil, 50mg/ml, was given. The final intraocular pressures were the same in both groups (mean of 12.6 mm Hg in the sequential group versus 12.2 mm Hg in the combined group, p=0.64).
In summary, except for the one study from Italy91, which differed from all other studies in the use of a two-site phacoemulsification procedure, no method of administration of 5-fluorouracil appeared to enhance the IOP lowering effect of combined cataract and glaucoma surgery.
Conclusion: 5-fluorouracil does not improve the IOP lowering effect of combined cataract and glaucoma surgery. [Evidence Grade: B]
Randomized controlled trials. Three randomized controlled trials of better than average overall quality (quality scores 64 percent95, 61 percent96 and 60 percent97) and one of lesser quality (quality score of 11 percent98) were reviewed.
In a German study, Borggrefe and others compared the results of single- versus two-site phacoemulsification and trabeculectomy, without anti-fibrosis agents, in eyes without prior incisional surgery.95 They found marked intraocular pressure lowering (mean of 13 to15 mm Hg) with greater than 1 year followup in both groups. The IOP reduction was 50 percent in the two-site group and 43 percent in the one-site group, a difference that was not statistically significant (p=0.262). There were no important differences in outcomes between the two groups.
In a Saudi Arabian study, El-Sayyad and colleagues randomized 76 eyes to either one- or two-site phacoemulsification and trabeculectomy, with the use of mitomycin-C.96 At 1 year, they found that both techniques effectively lowered IOP. The IOP lowering was slightly greater in the two- (mean 11.0 mm Hg) compared to the one-site (mean 7.1 mm Hg) group (no statistical analysis of IOP was provided by the authors), but the two-site group was also taking slightly more medication (mean 1.9 versus 1.6, p=0.6). Eighty-seven percent of the eyes in the two- and 84 percent in the one-site group had IOP less than or equal to 21 mm Hg without medications (p=0.47).
In a study from Italy, Rossetti and colleagues randomized 50 eyes to one- or two-site phacoemulsification and trabeculectomy, without the use of anti-fibrosis agents.98 IOP was lowered by a mean of 9.4 mm Hg in the one-site group and 11.0 mm Hg in the two-site group. Whether this small difference, which was not subjected to statistical analysis, was attributable to the operative technique or to baseline differences between the two groups cannot be determined.
In the United States, Wyse and colleagues randomized patients to single- versus two-site surgery using mitomycin-C 0.4 mg/ml for 2 minutes.97 With greater than 1 year followup, they found a slightly greater lowering of IOP (31 percent) in the two-site group than in the one-site group (23 percent, p=0.294), and a greater need for medications in the single-site group (mean of 0.9 medications per eye versus 0.2 medications per eye, p=0.03).
Non-randomized studies. For the non-randomized studies, quality scores tended to be low (24 percent54, 28 percent99, and 31 percent100). Bellucci and Luraschi performed two-site surgery in 10 patients and one-site surgery in 10 patients.54 The IOP lowering effect was similar in both groups, but the one-site group required more medication. The lack of randomization and the small number of patients precludes drawing firm conclusions from the data.
In a European population, Anand and colleagues compared three techniques: ECCE with nuclear expression and trabeculectomy through a single limbal site, ECCE with nuclear expression through clear cornea and trabeculectomy through a second site, and phacoemulsification and trabeculectomy through a single limbal site.99 No differences in IOP lowering were noted among the three groups (p=0.90 by analysis of variance).
Gayton and others compared 50 eyes undergoing one-site combined surgery with 65 patients undergoing two-site surgery.100 Mitomycin-C was used in some cases. The authors state that there were no significant differences at any time period in the amount of IOP lowering between the two groups.
In summary, although none of the four randomized controlled trials demonstrated a statistically significant difference in IOP lowering between one- and two-site surgery, they all showed a trend toward either lower IOP and/or reduced glaucoma medications in two-site surgeries.
Conclusion: Two-site combined trabeculectomy and cataract extraction may provide slightly lower IOP than one-site surgery. [Evidence Grade: C]
Randomized controlled trials. We found no randomized controlled trials addressing this question.
Non-randomized studies. Six studies reported a greater IOP reduction in eyes undergoing phacoemulsification combined with trabeculectomy than in eyes undergoing nuclear expression combined with trabeculectomy and three studies found no difference. The mean overall quality score of these 9 studies was 42 percent (quality scores 50 percent101, 46 percent102, 38 percent103, 43 percent104, 33 percent93, 54 percent105, 36 percent107, 28 percent99, and 50 percent106).
In Australia, Chia and Goldberg compared 50 eyes that underwent phacoemulsification and trabeculectomy with 50 that underwent ECCE with nuclear expression; 5-fluorouracil was used in both groups.101 With 1 to 2 years followup, the IOP lowering was greater (mean 5.8 mm Hg) in the phacoemulsification group than in the ECCE group (mean 2.5 mm Hg) and there were fewer complications. Shingleton and colleagues compared 35 eyes undergoing ECCE with nuclear expression and trabeculectomy with 37 eyes undergoing phacoemulsification and trabeculectomy.102 Phacoemulsification resulted in slightly greater IOP lowering (mean 5.4 mm Hg) than ECCE with nuclear expression (mean 3.2 mm Hg). Tezel and colleagues compared 127 eyes undergoing combined surgery with ECCE (nuclear expression) with 270 eyes undergoing combined surgery with phacoemulsification.103 Both groups received mitomycin-C, 0.2 mg/ml for 2 minutes. With 12 to 24 months followup, IOP was lowered a mean of 6.5 mm Hg in the ECCE group and 7.9 mm Hg in the phacoemulsification group. Wedrich and colleagues compared 54 eyes undergoing ECCE with nuclear expression and trabeculectomy with 49 eyes undergoing phacoemulsification with trabeculectomy.104 With a followup of at least 24 months, IOP was lower in the phacoemulsification group. Within a study primarily designed to evaluate the effect of 5-fluorouracil upon combined cataract and glaucoma surgery, Hurvitz compared the results of 55 eyes undergoing phacoemulsification and cataract surgery without anti-fibrosis agent with 36 eyes undergoing ECCE with nuclear expression and cataract surgery without anti-fibrosis agent.93 IOP was lowered a mean of 3.4 mm Hg in the nuclear expression group and 5.3 mm Hg in the phacoemulsification group. However, possible selection biases and the surgeon's learning curve could account for the greater effect observed in the phacoemulsification group. Wishart and Austin in England compared 34 eyes that underwent trabeculectomy and phacoemulsification with the previous 34 combined cataract and glaucoma operations that they performed by ECCE (nuclear expression).105 The IOP was lowered at 1 year by a mean of 8.6 mm Hg in the phacoemulsification group and by 5.2 mm Hg in the ECCE (nuclear expression group).
Three studies reported no difference in IOP reduction in eyes undergoing phacoemulsification combined with trabeculectomy compared to eyes undergoing nuclear expression combined with trabeculectomy. In a European population, Anand and others compared three techniques: ECCE with nuclear expression and trabeculectomy through a single limbal site, ECCE with nuclear expression through clear cornea and trabeculectomy, and phacoemulsification and trabeculectomy through a single limbal site.99 No differences in IOP lowering were noted among the three groups. Posterior capsular opacification requiring capsulotomy was less common in the phacoemulsification group. In England, Kosmin and others compared 32 eyes undergoing ECCE with nuclear expression and trabeculectomy with 31 eyes undergoing phacoemulsification and trabeculectomy.106 With greater than 2 years followup, there was little difference in the amount of IOP lowering or need for glaucoma medications. Stewart and colleagues compared the results of trabeculectomy and phacoemulsification in 16 eyes with trabeculectomy with ECCE (nuclear expression) in 14 eyes with followup of 1 year.107 No significant differences were found between the two techniques, although the study was very small.
Manoj and colleagues, in England, studied a related question: whether ECCE with nuclear expression or phacoemulsification provided better IOP control when cataract extraction was performed in eyes with prior trabeculectomy.47 Phacoemulsification appeared to be better than nuclear expression in terms of IOP and need for IOP lowering medications at 2 years, but possible baseline differences between the two groups complicates interpretation of the results.
Honjo and colleagues, performed a unique study comparing three methods of combining cataract extraction with trabeculotomy (rather than trabeculectomy).50 Twenty-five eyes underwent nuclear expression, 10 eyes single-site surgery with phacoemulsification, and 12 eyes two-site surgery with phacoemulsification. The nuclear expression group had a mean IOP lowering of 9.6 mm Hg on 1.1 medications; the two-site phacoemulsification lowered IOP by a mean of 7.4 mm Hg on 0.6 medications, and the one-site phacoemulsification lowered IOP by 11.6 mm Hg on 1.4 medications. Thus there appeared to be little difference between the techniques. However, the baseline characteristics of the groups were not well characterized, and the number of subjects was small.
In summary, although no randomized controlled trials on this topic were performed, the preponderance of studies find that IOP is lower in eyes undergoing phacoemulsification rather than nuclear expression when combined with trabeculectomy.
Conclusion: IOP is lowered more when phacoemulsification rather than nuclear expression is used in combined cataract and glaucoma surgery. [Evidence Grade: C]
Randomized controlled trials. We found no randomized trials addressing this issue.
Non-randomized studies. Quality scores were calculated for only one of the non-randomized studies addressing this question. Donoso and Rodriguez compared the results of single incision, combined trabeculectomy and phacoemulsification with that of a matched group of eyes undergoing phacoemulsification after trabeculectomy.94 In both groups, a single intraoperative application of 5-fluorouracil, 50 mg/ml, was given. The final intraocular pressures were the same in both groups.
El-Sayyad and colleagues matched 53 consecutive eyes undergoing simultaneous phacoemulsification and trabeculectomy with mitomycin-C through a single site with 53 eyes that underwent trabeculectomy with mitomycin-C followed at a later date by phacoemulsification.86 In the latter group, the phacoemulsification was through clear cornea in 32 eyes and at the limbus in 21 eyes. Both protocols yielded similar and marked IOP reductions.
Murchison and Shields compared three surgical approaches in patients with coexisting cataract and glaucoma.108 Two of these approaches, simultaneous glaucoma and cataract surgery, and filtering surgery followed by cataract surgery, bear directly on the question of interest. Twenty-two eyes with compelling reasons to undergo simultaneous cataract and glaucoma surgery, that underwent a combined guarded filtering procedure, were compared to 22 eyes whose most pressing problem was IOP control requiring filtration surgery, that underwent first filtration surgery alone, and then cataract surgery at a later date. The cataract surgery technique was ECCE with nuclear expression; the filtering procedure was either trabeculectomy or a full thickness procedure. The final IOP at 12 to 24 months was similar in the two groups, although the eyes undergoing the staged procedure had much higher IOPs to begin with. Only 14 percent of eyes undergoing the combined procedure had an increase in IOP in the first 2 postoperative days compared to preoperatively. This study is difficult to interpret because the two groups differed so markedly in their characteristics.
In summary, three studies, none of them randomized, directly compared staged versus combined procedures. Within this limited database, the overall outcome of combined cataract and glaucoma surgery, whether employing nuclear expression or phacoemulsification, and regardless of anti-fibrosis regimen, appears to be similar for staged and combined procedures.
Caprioli and colleagues, carefully matched 40 consecutive eyes undergoing temporal corneal phacoemulsification following trabeculectomy with 40 eyes that had undergone trabeculectomy but did not have subsequent cataract surgery.108 The survival curves for IOP control were the same for both groups of eyes, suggesting that the cataract surgery did not adversely affect IOP. Although not directly addressing the question of whether combined procedures or staged procedures are preferable, the inference is that IOP can be effectively lowered by a staged procedure.
Conclusion: There is insufficient evidence to conclude that staged or combined procedures give better results. [Evidence Grade: I]
Randomized controlled trials. Gayton and colleagues (study quality score 51 percent), performed a randomized controlled trial of trabeculectomy versus endoscopic laser cycloablation, each combined with cataract surgery, in 58 eyes.110 The combined procedures using trabeculectomy used a two-site approach, and some eyes received mitomycin-C. With a mean followup of 2 years, eyes receiving trabeculectomy had a greater reduction of IOP (mean 10.6 mm Hg) than those receiving endolaser (mean 7.3 mm Hg).
Gianoli, Schnyder, Bovey et al. (study quality score 59) randomized 60 eyes of 60 patients to receive either combined phacoemulsification and deep sclerectomy or combined phacoemulsification and trabeculectomy.111 The former procedure was performed as a two-site operation, the latter as single-site. At 12 months the IOP was lowered by 10.9 mm Hg in the deep sclerectomy eyes and by 9.89 mm Hg in the trabeculectomy eyes, a non-significant difference. The IOPs at entrance were similar in the two groups (24.8 mm Hg in the deep sclerectomy group and 24.6 mm Hg in the trabeculectomy group) as were the number of post-operative glaucoma medications (0.6 in both groups). It should be noted that data on the number of eyes remaining in the study at the postoperative time points was not provided and the two groups differed not only in glaucoma procedure but in one- versus two-sites.
Non-randomized studies. Kendrick and others (quality score 20), reported a cohort study comparing ab interno laser thermal sclerostomy combined with cataract surgery versus trabeculectomy combined with cataract surgery.112 At 6 months they found a mean IOP lowering of 5.2 mm Hg in the laser thermal sclerostomy group and 2.3 mm Hg in the trabeculectomy group. Problems with comparability of the two groups limited the validity of the results.
In summary, few trials of other techniques besides trabeculectomy have been reported in combined cataract and glaucoma operations. One study does suggest comparable IOP lowering effects of deep sclerectomy and trabeculectomy when combined with phacoemulsification. Further research will be necessary before other glaucoma procedures replace trabeculectomy in combined cataract and glaucoma operations.
Conclusion: There is insufficient evidence to compare the effectiveness of alternative glaucoma procedures to trabeculectomy in combined procedures. [Evidence Grade: I]
Of the 110 reports of trials included in the review, only 92 studies reported data on complications. The complications were divided into those that occur soon after surgery (early) and late complications, defined typically as more than 1 month after surgery.
Rupture of the posterior capsule of the lens is an intraoperative complication that makes implantation of an artificial intraocular lens more difficult and increases the probability of other complications, some of which are vision-threatening. Overall, the rates of capsule rupture in these reports were modest and not dramatically different with rates for persons undergoing cataract surgery who do not have glaucoma, averaging less than five percent of eyes. Two reports, Berestka and Brown15, and Murchison and Shields17 documented substantially higher capsule rupture in eyes undergoing combined trabeculectomy/lens implant surgery when a limbus-based conjunctival flap was used (18 percent in each series) compared to eyes in these series that had fornix-based flaps.
Early leakage of the wound in the area of a trabeculectomy bleb occurred in zero to 10 percent of operations in a variety of settings. Some reports documented higher rates when certain surgical techniques were used. Wong and colleagues found 25 percent leakage when 5-fluorouracil was used in conjunction with cataract/trabeculectomy surgery compared to eight percent leakage without 5-fluorouracil.89 Cohen found 5-fluorouracil usage associated with 59 percent early leakage compared to 19 percent without it.92 It is intuitively consistent that anti-fibrosis treatment (5-fluorouracil) would slow healing and lead to early leakage.
Further study of early leakage by Chia and Goldberg noted the striking rate of 44 percent early leaks in eyes having extracapsular cataract surgery compared to 24 percent leakage in those undergoing phacoemulsification.101 Other differences in leakage were reported for fornix-based flaps (eight percent leakage) versus limbus based (less than one percent), by Tezel and colleagues16 and for two-site cataract/trabeculectomy surgery (9 percent) compared to one-site surgery (4 percent), by El Sayyad and colleagues.86 It is safe to say that reporting criteria for such leakage must have varied dramatically (and in ways that are not apparent from the reports themselves) to lead to such large differences in absolute rates.
Choroidal effusion is a finding associated with lower eye pressure in the postoperative period, and a complication that is notoriously under-reported unless examinations are specifically targeted to find it. In the 89 reports evaluated, choroidal effusion is noted in 47, with up to 15 percent of eyes having this complication. Among eyes that underwent combined cataract/trabeculectomy surgery, there were only a few associations with higher effusion rates. Fornix-based conjunctival flaps were reported to be associated with choroidal effusion in eight percent of eyes compared to two percent of eyes having limbus-based flaps.14 Reports by Kosmin21 and by Lyle18 reached contradictory conclusions regarding the association of effusions and incision size in combined cataract trabeculectomy surgery. The former reported effusions in 15 percent of those with smaller incisions compared to three percent of those with larger incisions. The latter reported effusions in 12 percent of those with larger incisions compared to three percent of those with smaller incisions.
Early endophthalmitis was essentially unreported, and since this complication is highly evident and typically a dramatic event, it is likely that few reports actually had cases of this problem in their populations studied. However, the likely rate of one infection per 1,000 operations would never have been detected in studies with the small sample sizes evaluated here.113
Epithelial defects are a problem largely engendered by the use of anti-fibrosis treatment with 5-fluorouracil. In eyes undergoing combined phacoemulsification and cataract extraction, O'Grady and colleagues reported 10 eyes with epithelial defects in the 5-fluorouracil treated eyes versus zero eyes in the control group.90 Wong and colleagues 89 reported an incidence of 67 percent in eyes receiving 5-fluorouracil versus 35 percent in the controls; whereas Chia101 noted a rate of 56 percent among extracapsular cataract procedures combined with trabeculectomy compared to 32 percent in phacoemulsification trabeculectomy, both groups being treated with 5-fluorouracil. None of the studies using mitomycin-C for anti-fibrosis reported any epithelial effects-consistent with the lack of effect of this drug on epithelial cells.
A severe (fibrinous) inflammatory reaction occurred in one percent or fewer of eyes in most reports, though lack of comment may not necessarily mean that no such reactions occurred. However, some investigations reported high rates of fibrinous reactions, including a study by Chia101 using 5-fluorouracil, a study by Cohen79 using mitomycin-C, and a study by Borggrefe95 not using anti-fibrosis agents. However, the rates of severe inflammation in these reports varied from 20 percent to 60 percent. By contrast, Krupin and colleagues found no examples of fibrinous reactions among 68 cases of extracapsular surgery.51 The definition of fibrinous reaction may have varied dramatically among these studies and may explain the inconsistent conclusions.
Hyphema or bleeding into the anterior chamber soon after surgery can vary from barely detectable to a large amount of blood. Again, for this early complication, rates reported vary from less than 10 percent of eyes undergoing trabeculectomy alone or combined with cataract surgery in the largest series, to more than 50 percent of eyes in one report. Reports indicated possibly higher rates with two-site combined surgery100, higher doses of mitomycin-C72, phacoemulsification compared to nuclear expression101, extracapsular surgery105, and higher in large compared to small incision cataract/ trabeculectomy incisions.18
Hypotony can lead to secondary vision loss on a short-term or permanent basis. The typical rate among many studies reviewed for low pressure (usually considered to be 5 mm Hg or less) is 10 to15 percent. Blondeau and Phelps reported a higher rate of hypotony for thermosclerostomy versus trabeculectomy glaucoma surgery (20 percent versus 10 percent).74 Borggrefe and others found hypotony in 16 percent of eyes undergoing a one-site combined procedure compared to 4 percent with a two-site cataract/trabeculectomy.95 Lemon and others83 found a higher rate of hypotony with limbus versus fornix flaps (30 percent versus eight percent) and Wishart and Austin105 reported hypotony in 36 percent of eyes after nuclear expression versus 15 percent of eyes after phacoemulsification.
Likewise, suprachoroidal hemorrhage is a vision threatening condition related to hypotony and wound leaks. This complication was highest in a series by Sanders and others72 using mitomycin-C (six percent), and lowest in combined cataract/trabeculectomy both with an anti-fibrosis agent (one percent)77 or without an anti-fibrosis agent (one percent).16, 103 This would be expected if combined procedures give fewer low pressures due to more complete healing of the surgery.
The final early complication documented was vitreous hemorrhage, seen infrequently, and at its maximum (six percent) in eyes undergoing trabeculectomy with mitomycin-C treatment.72 In combined procedures using mitomycin-C a representative rate would seem to be that in a study by Tezel's group, who reported eight cases of vitreous hemorrhage in 397 eyes (two percent).103 This complication is unlikely to have escaped detection, as it is a cause of serious vision decrease and is readily visible.
Late complications included the Tenon cyst, a solid elevation of the bleb usually associated with higher than desired eye pressure. This was not often noted in most studies, although whether due to lack of occurrence or failure to report cannot be determined. The highest rate was three percent in one trabeculectomy series by Cheung and colleagues.62
Late bleb leakage was also infrequently reported. Of 300 eyes undergoing surgery in a controlled trial in India, Robin and colleagues, reported one leakage during 12 months followup.70 Shingleton14 and Caprioli 23 respectively had less than one percent, and zero of 80 cases, respectively, of this problem to report. The problem of late wound leakage (any leakage related to the wound) is easily confused with late bleb leakage, (leaks confined to the surface of the bleb). It is not clear that each study differentiated these in this way. In one of the few long-term studies, Lewis and Phelps found that 13 percent of eyes after glaucoma surgery alone developed a wound leak (almost surely this is identical to a bleb leak).78 In combined surgery, leakage was less common (three percent at 1 year after combined surgery without anti-fibrosis, as reported by Anand and colleagues99), and less than one percent of 612 eyes in combined surgery with mitomycin-C as reported in two studies by Tezel's group.16, 103
Cystoid macular edema was noted in fewer than two percent of cases in most studies, and was remarkable for the lack of differences related to technique variations. This is possibly due to the lack of performance of fluorescein angiography needed to make any definitive statement about the presence of this complication.
Late endophthalmitis was noted in only 3 eyes of all the reports, Carlson80, Sanders72, and Yang.27 Since it occurs at a rate of fewer than one per 1,000 per year of followup, each series would have only trivial power to detect this problem.113
Corneal graft failure was not noted in any study, surely due to the lack of significant numbers of cases of those who had penetrating keratoplasty with secondary glaucoma in these reports.
Hypotony maculopathy was not reported in any of the series that did not use an anti-fibrosis agent. Hence, it can be speculated that it is either a direct result of the drugs themselves or a result of the low pressure that is generated by their use. Robin70 found three percent of trabeculectomies with mitomycin-C treatment had hypotony maculopathy, while none of those without mitomycin-C had it. Budenz found one of 71 eyes with this complication, and it did not occur in the group not receiving anti-fibrosis agents.84 Tezel's group16, 103 and El- Sayyad and colleagues86 detected it in eight percent and two percent, respectively, of their combined cataract/trabeculectomy series, all of whom had received mitomycin-C.
Posterior capsule opacity leads to decreased vision, but there is difficulty in determining whether authors were reporting simply its initial presence or an opacity sufficient to obscure vision or to lead to YAG laser capsulotomy to relieve the problem. Furthermore, the progressive opacity should be evaluated as an incident rate per unit time, and many of our studies did not report the length of followup or if they did, the timing of the capsule opacity was not given. Within the first 2 years after cataract surgery in glaucoma subjects, between 12 percent27 to 50 percent101, of eyes were reported to develop capsule opacity. The rate did not appear associated with any particular method of surgery, nor was it related to use of anti-fibrosis therapy. Two large cohort studies by Tezel and colleagues16, 103 of phacoemulsification combined surgery had a rate of 25 percent, while a large series of extracapsular surgery42, had a similar rate, 30 percent. Clearly, longer followup would be likely to increase the proportion affected and some method of standardizing the reporting of this problem would be helpful.
Shin, Kim, Ren and colleagues87 evaluated the incidence of posterior capsular opacification in eyes without previous glaucoma surgery receiving mitomycin-C or placebo at the time of combined phacoemulsification and trabeculectomy. Using life table analysis they found that YAG capsulotomy was performed less frequently in the eyes that received mitomycin-C (p = 0.004). For instance, the proportion of eyes receiving capsulotomy by 4 years after surgery was approximately 60 percent in the placebo group and 40 percent of the mitomycin-C group.
Glaucoma surgery was associated with an increased risk of postoperative cataract.
A glaucoma procedure added to cataract surgery lowers IOP more than cataract surgery alone.
Limbus- and fornix-based conjunctival incisions provided the same degree of long-term IOP-lowering in combined surgery.
In combined surgery using phacoemulsification, the size of the cataract incision did not affect long-term IOP control.
5-fluorouracil was not beneficial in further lowering IOP when used with combined procedures.
Mitomycin-C was efficacious in producing lower long-term IOPs when used with combined procedures.
Combined procedures resulted in lower IOP at 24 hours than cataract extraction alone.
ECCE alone appears to increase IOP at 24 hours.
In the long-term, cataract surgery alone lowered IOP by 2 to 4 mm Hg, combined cataract and glaucoma surgery lowered IOP by 6 to 8 mm Hg, and the performance of a glaucoma procedure alone provided even greater long-term IOP-lowering than combined cataract and glaucoma surgery.
Combined surgery in which the incisions for the cataract extraction and glaucoma procedure are separate provided slightly lower long-term IOP than a one-site approach.
Combined surgery in which phacoemulsification is used provided slightly lower long-term IOP than nuclear expression.
The same complications occurred with combined surgery as occur with cataract and glaucoma surgery performed separately.
5-fluorouracil use was associated with epithelial defects.
Hypotony was more likely after the use of anti-fibrosis agents than after surgery performed without antifibrosis agents.
Although the literature on the management of coexistent cataract and glaucoma allowed us to draw the above conclusions, it had important limitations.
First, in terms of the management of glaucoma, we were forced to limit our review to the effect of surgery upon intraocular pressure (IOP). It is well accepted that the level of IOP is not the disease, but simply an important risk factor. The effect of surgery upon the clinical course of glaucoma would be more appropriately determined by evaluating its long-term effect upon the appearance of the optic nerve, visual field, and quality of life. Very few of the articles that we reviewed reported data on the optic nerve or visual field.
The reasons for this deficiency in the literature are several. First, glaucoma is generally a slowly progressive disease, so to detect worsening or stabilization of the optic nerve or visual field would take a study of at least 3, and more likely 5 to 10 years, in duration. Most of the articles reviewed had much shorter durations of followup. Secondly, because of the variability of visual field testing, it is a daunting task, even over a several year period, to definitively document deterioration of the visual field. Likewise, our ability to detect changes in the appearance of the optic nerve in a reproducible fashion is limited. Finally, the presence of coexisting cataract and glaucoma presents unique problems in monitoring the optic nerve and visual field. Specifically, the presence of a cataract can make preoperative assessment of the optic nerve difficult or impossible, because the examiner, or examining device, cannot see the optic nerve through an opaque cataract. Furthermore, cataract and its removal affect the visual field, and it can be difficult to separate out whether changes in the visual field are due to cataract or glaucoma. In an eye with increasing cataract after glaucoma surgery, it is particularly difficult to determine whether worsening in the visual field is due to glaucoma or cataract.
In terms of cataract, the literature was generally deficient in any objective description of the ocular lens, whether clear or cataractous. Many studies simply state that the eyes undergoing surgery had cataracts, without providing an assessment of the severity of the cataract. Particularly in the studies commenting on cataract after glaucoma surgery, we usually could not tell if cataract was simply an observed lens opacity, or visually significant. No standardized grading of cataract was reported.
Assessment of the literature was limited by its quality. Although the overall study quality score for the articles reviewed was higher (53 percent) than in some other evidence reports, such as the Report on Anesthesia Management During Cataract Surgery (overall mean study quality score of 44 percent), in some categories of study quality, particularly bias and confounding the score was very low (mean of 30 percent).6 In many studies, a lack of comparability of study groups and an absence of masking of observers greatly compromised their validity.
In addition to these general comments concerning deficiencies in the literature we uncovered gaps in the literature concerning the specific questions that were asked.
There were only two randomized controlled trials comparing phacoemulsification to phacoemulsification plus trabeculectomy. One of those trials enrolled only 10 patients per group. The other trial used an unusual approach to performing the trabeculectomy (a no-stitch technique). While the results of these two studies agreed with the expectation that combined procedures produce a larger decrease in IOP than cataract extraction alone, the evidence is weak.
The evidence is weak that phacoemulsification plus trabeculectomy results in less IOP lowering than trabeculectomy alone. No randomized trials addressed this question. This is an important issue when deciding whether or not to perform a combined procedure in a glaucoma patient with uncontrolled glaucoma.
While extracapsular cataract extraction (ECCE) and phacoemulsification both appear to lower IOP, the evidence that phacoemulsification lowers IOP by a greater amount was only found when comparing the results of case series and non-randomized controlled trials. If indeed phacoemulsification reduces IOP more than ECCE, then clinicians might alter the surgical approach in individuals with ocular hypertension or glaucoma. A randomized controlled trial would help to address this question.
The evidence that phacoemulsification does not harm functioning filtering blebs is quite limited, with only two articles looking at this question. Further research in this area would be of great importance to clinicians. If the IOP lowering effect is greater performing trabeculectomy alone than in a combined procedure, and if the bleb is not affected by later clear cornea phacoemulsification, then this would argue for sequential surgery in most patients with cataract and glaucoma.
While early studies found large increases in IOP immediately after ECCE in patients with glaucoma, later research with phacoemulsification found mean decreases in IOP on the first postoperative visit. Data are limited, however, and larger series are needed to confirm that immediate IOP spikes are not common using phacoemulsification techniques. If immediate IOP spikes are not common after phacoemulsification alone in glaucoma eyes, then it might not be necessary to perform combined surgery as often.
The data supporting different regimens to control postoperative IOP spikes in individuals with glaucoma are limited. More research on the best regimen for preventing immediate postoperative IOP spikes would be valuable to clinicians.
Several studies have been performed, both with and without anti-fibrosis agents. The consensus is that two-site surgery may result in slightly lower IOP in aggregate than single-site surgery. However, none of the studies included African-Americans, a group with a high prevalence of glaucoma and in whom responses to surgery may differ from Caucasians.
The number of studies (three) addressing this question was small, and no randomized trial has been performed. Even if the ultimate effect on IOP is better with the two-staged procedure, quality of life concerns about the disadvantages of deferring visual improvement and the necessity of two operations in the staged strategy need to be addressed.
Only one study looked at the role of mitomycin-C in African-American patients. No consensus exists in the literature regarding the optimal dose and duration of mitomycin-C use. Furthermore, no studies have attempted to confirm or refute the interesting finding of decreased posterior capsular opacification in eyes treated with mitomycin-C. The finding in one study of decreased posterior capsular opacification is an intriguing finding that merits further investigation.
The literature is scant on the use of endoscopic laser treatment and non-penetrating glaucoma surgery in conjunction with cataract surgery. Before other procedures are adopted, more and higher quality studies will need to be done.
The findings of this report can serve both to give support to certain strategies now being used to manage coexistent cataract and glaucoma and to highlight areas in which physician behavior is not informed by adequate information. It can serve as an impetus to future research on the important unresolved questions and point toward better ways of designing and reporting the results of future studies. Chapter 5 contains a more detailed discussion of future research implications.
As discussed in the conclusions, the literature addressing management of coexistent cataract and glaucoma has important limitations. We will first address those limitations general to the entire literature and then those that are specific to individual questions.
As mentioned previously, although glaucoma is characterized by structural and functional damage to the optic nerve, the extant literature concentrates on intraocular pressure (IOP). It will be important for future studies to assess the effect of surgery upon the appearance of the optic nerve, visual field, patient preferences and quality of life in addition to IOP. Because of the chronic and slowly progressive nature of glaucoma, these parameters need to be evaluated in long-term (5 years or greater) studies.
Future studies should pay more attention to the quantification of cataract and its effect upon vision. The existing literature is generally lacking in any objective description of the crystalline lens, whether clear or cataractous. Quantifying lens opacity and assessing its effect upon vision and vision related quality of life should be a priority for future studies. Standardized grading systems for cataract (already in use) should be employed.
Although the literature represented populations in North America, Europe, and Asia, few of the reports included significant numbers of African-Americans. This population has a high prevalence of glaucoma, and may respond differently to glaucoma surgery. Future researchers must strive to include more African-Americans in their studies.
As described in Chapter 4: Conclusions, assessment of the evidence was limited by its quality, particularly in the areas of bias and confounding. Specifically, a lack of comparability of study groups and an absence of masking of observers greatly compromised the validity of many studies. Future studies would be well advised to pay particular attention to striving for comparability of study groups and statistical adjustment for baseline differences, as well as implementing more objective measures for assessing outcomes. A large number of ophthalmology journals are signatories to the CONSORT statement, which outlines standardized criteria for reporting the results of randomized clinical trials.114 Adherence to these standards is one way that trials reporting may be improved.
The following specific questions point to potential areas for future research, both because of their importance and the fact that questions remain unresolved by the extant literature:
Does trabeculectomy alone lower IOP more than phacoemulsification plus trabeculectomy?
Does phacoemulsification harm functioning filtering blebs?
Does phacoemulsification reliably result in lower IOP on the first day after surgery? If so, then it might not be necessary to perform combined surgery.
Does a two-stage procedure provide lower IOP than a combined procedure?
What is the role of mitomycin-C in combined procedures and how should it be administered?
Is the finding of decreased posterior capsular opacification with the use of mitomycin-C reproducible?
Can better glaucoma procedures be developed to be used in conjunction with cataract surgery?
What are the risk factors for cataract development and progression following glaucoma surgery?
| 5-FU | 5-fluorouracil |
| AC | Anterior chamber |
| ACG | Angle closure glaucoma |
| ALT | Argon laser trabeculoplasty |
| ATT | Argon laser trabeculoplasty - trabeculectomy - trabeculectomy treatment sequence |
| BSS | Balanced salt solution |
| CE | Cataract extraction |
| CME | Cystoid macular edema |
| DSCI | Deep sclerectomy with collagen implant |
| ECCE | Extracapsular cataract extraction |
| Epi. | Corneal epithelial |
| Hem. | Hemorrhage |
| ICCE | Intracapsular cataract extraction |
| IOL | Intraocular lens |
| IOP | Intraocular pressure |
| MMC | Mitomycin-C |
| N/A | Not applicable |
| No CE | No cataract extraction |
| NR | Not reported |
| NRCT | Non-randomized controlled trial |
| OAG | Open-angle glaucoma |
| PC | Posterior capsule |
| PCO | Posterior capsular opacification |
| PECE | Phacoemulsification cataract extraction |
| Phaco. | Phacoemulsification |
| PMMA | poly(methyl methacrylate) intraocular lens |
| POAG | Primary open-angle glaucoma |
| RCT | Randomized controlled trial |
| SPK | Superficial punctate keratitis |
| TAT | Trabeculectomy - argon laser trabeculoplasty - trabeculectomy treatment sequence |
| Trab. | Trabeculectomy |
| VA | Visual acuity |
| YAG | YAG laser |
American Board of Eye Surgery
Robert Dotson, MD
American Glaucoma Society
Robert N. Weinreb, MD
American Optometric Association
John F. Amos, OD, MS
American Society of Cataract & Refractive Surgery
Alan Robin, MD
Cochrane Collaboration: Eyes & Vision CRG
Richard Wormald, MA, MB, MSc
Glaucoma Patient Perspective
Ms. Edith Marks
Glaucoma Research Foundation
Lars Vistnes, MD (patient)
Ruth Williams, MD
Health Care Financing Administration
John J. Whyte, MD, MPH
The Glaucoma Foundation
Ms. Ali Hodin
The Johns Hopkins University Evidence-based Practice Center will be developing an Evidence Report on Treatment of Coexisting Glaucoma and Cataract for the Agency for Healthcare Research and Quality of the U.S. Department of Health and Human Services. We are soliciting input from experts with varied perspectives on key questions regarding management of coexisting cataract and glaucoma.
PLEASE ASSIST US IN IDENTIFYING, FROM YOUR PERSPECTIVE AS _________________, WHAT ARE THE MOST IMPORTANT QUESTIONS THAT SHOULD BE ADDRESSED.
Please return your form to Dr. Lisa Lubomski at (410) 614-9651 (fax), by December 13th. Should you have any questions, please call us at (410) 614-4037 or e-mail (llubomski@jhmi.edu).
Please read the following eight questions regarding treatment of coexisting glaucoma and cataract and rank each question below as follows: (1) crucial, (2) desirable, or (3) unimportant, for in-depth examination in our study. Please write in the space provided any other questions that you think are particularly relevant or controversial.
| Possible Questions | Rank | Comments |
|---|---|---|
| 1. What is the effect of glaucoma surgery (i.e., filtration surgery) upon the development and progression of cataract? | ||
| 2. What is the effect of cataract surgery upon short and long-term intraocular pressure control in open angle glaucoma patients? | ||
| 3. What are the risks and benefits of the alternative surgical techniques for operating on eyes with both cataract and glaucoma: | ||
 a) Staged procedure versus combined? | ||
| b) With anti-fibrosis agent or without anti-fibrosis agent? | ||
| c) Single site for both operations or two sites? | ||
| d) Trabeculectomy versus endoscopic laser versus deep sclerectomy/viscocanulostomy? | ||
| e) Nuclear expression or phacoemulsification? | ||
| (OTHER) | ||
How might your organization use an evidence report on Treatment of Coexisting Glaucoma and Cataract? Check all that apply
Would use report to facilitate development of a new practice guideline on surgical treatment of patients with coexisting glaucoma and cataract.
Would use report to assess the appropriateness of your organization's existing guidelines on management of patients with coexisting glaucoma and cataract.
Would use report to prepare educational materials for physicians.
Would use report to prepare educational materials for patients.
Other (please specify)____________________________________
Other (please specify)____________________________________
Other Comments?_____________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
________
Signature of person completing form: __________________________________________
((((((“cataract”[mh] OR cataract[tw]) AND (“glaucoma”[mh] OR glaucoma[tw])) AND (((((((((“surgery”[mh] OR surgery[tw]) OR “filtering surgery”[mh]) OR filtering surgery[tw]) OR “cataract extraction”[mh]) OR cataract extraction[tw]) OR sclerostomy[tw]) OR trabeculectomy[tw]) OR phacoemulsification[tw]) NOT ((“child”[mh] OR “anesthesia”[mh]) OR “anesthetics”[mh])))
limited to English [la]
added to all phases of PubMed optimal search strategy for retrieval of controlled trials, with addition of search strategy for case series
(randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized controlled trials [mh] OR random allocation [mh] OR double-blind method [mh] OR single-blind method [mh] OR clinical trial [pt] OR clinical trials [mh] OR (clinic* [tw] AND trial* [tw]) OR ((singl* [tw] OR doubl* [tw] OR trebl* [tw] OR tripl* [tw]) AND (mask* [tw] OR blind* [tw])) OR (latin [tw] AND square [tw]) OR placebos [mh] OR placebo* [tw] OR random* [tw] OR research design [mh:noexp] OR comparative study [mh] OR evaluation studies [mh] OR follow-up studies [mh] OR prospective studies [mh] OR cross-over studies [mh] OR control* [tw] OR prospectiv* [tw] OR volunteer* [tw] OR case report[mh] OR (case[tw] AND series[tw])) NOT (animal [mh] NOT human [mh])
note: items underlined added to strategy for the glaucoma EPC literature search
| Journal Name | Cochrane Handsearching (dates of registered searches) a | PubMed (date, volume and issue of most recent citation) |
|---|---|---|
| American Journal of Ophthalmology | 1970, 1981, 1985-1998 | Dec 1999 128(6) |
| Archives of Ophthalmology | 1964-1998 | Jan 2000 118(1) |
| Journal of Cataract and Refractive Surgery | 1993-1995 | Dec 1999 25(12) |
| Journal of Glaucoma | 1992-present | Dec 1999 8(6) |
| Ophthalmology | 1970, 1978-1998 | Dec 1999 106(12) |
Note: In addition to handsearching, the Cochrane Collaboration also searches electronic databases, such as MEDLINE and EMBASE, for controlled trials.
| Question | Evidence Table |
|---|---|
| What are the risks and benefits of staged versus combined procedures? | 10 |
| What is the effect of cataract surgery on short-term IOP control in open angle glaucoma? | 5 |
| What is the effect of cataract surgery on long-term IOP control in open angle glaucoma? | 4 |
| What is the effect of glaucoma surgery on the development and progression of cataract? | 6 |
| What are the risks and benefits of mitomycin-C in combined cataract and glaucoma surgery? | 7 |
| What are the risks and benefits of 5-fluorouracil in combined cataract and glaucoma surgery? | 8 |
| What are the risks and benefits of trabeculectomy combined with cataract surgery compared with other types of glaucoma surgery combined with cataract surgery? | 11 |
| What are the risks and benefits of single-site versus two-site operations? | 9 |
| What are the risks and benefits of nuclear expression versus phacoemulsification in combined cataract and glaucoma surgery? | 9 |
| TOPIC | EVIDENCE GRADES | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| EB | NC | DF | JK | HJ | LL | HQ | FINAL GRADE | PAGE | |
| CE alone versus CE and glaucoma procedure (long-term) | B+ | A | B | A | A | A | B | A | 36 |
| Limbus- versus fornix-based combined procedures | B | B | B | B | B | B | B | B | 37 |
| Small versus large incision PECE | B | B | B | B | B | B | C (B) | B | 39 |
| PECE plus trabeculectomy versus trabeculectomy alone | C | C | C | B | B | C | C | C | 40 |
| PECE plus trabeculectomy | C | B | B | C | B | C | B | C | 41 |
| ECCE plus trabeculectomy versus trabeculectomy | B | B | C | B | B | C | C | B | 42 |
| ECCE plus trabeculectomy | C | C | B | C | B | C | C | C | 43 |
| ECCE alone | C | B | B | C | B | C | B | C | 44 |
| PECE alone | C | B | B | C | B | C | B | C | 45 |
| CE with functioning filtering bleb | I | C | B | B | I | C | I | 47 | |
| CE alone versus CE plus glaucoma procedure (short-term) | B | B | C | B | B | B | I | I | 49 |
| Adjuvant medications in combined ECCE and trabeculectomy | B | C | C | C | C | C | C | C | 50 |
| PECE combined procedures | C | B | B | C | C | B | I | C | 51 |
| PECE alone | I | I | I | I | I | 51 | |||
| ECCE combined procedures | I | B | C | C | C | C | B | C | 52 |
| ECCE alone | C | C | C | C | C | 52 | |||
| Effect of glaucoma surgery on cataract | A- | A | B | A | A | B | A | A | 55 |
| Effect of mitomycin-C on IOP in combined surgery | C | C | I | B | C | B | B | 60 | |
| Effect of 5-FU on IOP in combined surgery | B | A | B | A | A | B | B | B | 63 |
| One- versus two-site procedures | B | C | B | A | B | C | C | 66 | |
| Nuclear expression versus phacoemulsification | C | B | B | C | B | C | I | C | 69 |
| Staged versus combined procedures | I | B | I | I | I | I | C | I | 71 |
| Trabeculectomy versus other procedures | I | B | I | I | I | I | I | I | 73 |
| Number of Citations | |||||
|---|---|---|---|---|---|
| Source | Search Strategy | Date | Retrieved | Unique (included in abstract review process) | Included in Article Review Process |
| PubMed | PubMed core | Nov 22, 1999 | 781 | 781 | 120 |
| CENTRAL | CENTRAL core | Nov 22, 1999 | 85 | 6 | 3 |
| Issue 4 1999 | |||||
| CENTRAL | CENTRAL core | Feb 14, 2000 | 88 | 3 | 0 |
| Issue 1 2000 | |||||
| PubMed | PubMed core | Mar 13, 2000 | 795 | 18 | 4 |
| Handsearch | table of contents of priority journals | to Apr 15, 2000 | 6 | 3 | 1 |
| PubMed | revised PubMed (citations uniquely retrieved by change to strategy) | Jul 18, 2000 | 109 | 108 | 3 |
| TOTAL | 1864 | 919 | 131 | ||
| Exclusion Criterion on Abstract Review Forma | Number Excluded |
|---|---|
| Did not include human data | 13 |
| Adults not part of study population | 0 |
| Not in English | 1 |
| No original data | 28 |
| Does not address open angle glaucoma or primary open angle closure glaucoma | 483 |
| Was neither a controlled trial nor a case series ≥ 100 patients | 403 |
| Only addresses intracapsular cataract extraction | 27 |
| Only addresses full thickness glaucoma surgery | 2 |
| Meeting abstract | 0 |
| Other (e.g., does not address cataract surgery) | 30 |
Note: Reviewers did not have to agree on reason for exclusion
| Study Topic on Abstract Review Forma | Number of Abstracts |
|---|---|
| Development of cataract after glaucoma surgery | 32 |
| Acute control of IOP after cataract surgery, in glaucoma patients | 17 |
| Long-term control of IOP after cataract surgery, in glaucoma patients | 45 |
| Preferred surgical technique | 75 |
Note: Reviewers did not have to agree on topic and could select multiple topics for each abstract.
Bobrow JC. Prospective intrapatient comparison of extracapsular cataract extraction and lens implantation with and without trabeculectomy. Am J Ophthalmol 2000;129:291-6.
Collaborative Normal-Tension Glaucoma Study Group. Comparison of glaucomatous progression between untreated patients with normal-tension glaucoma and patients with therapeutically reduced intraocular pressures. Am J Ophthalmol 2000;126(4):487-97.
Dreyer EB, Chaturvedi N, Zurakowski D. Effect of mitomycin C and fluorouracil-supplemented trabeculectomies on the anterior segment. Arch Ophthalmol 2000;113(5):578-80.
Fanous S, Brouillette G. Combined trabeculectomy and cataract extraction: modified technique. Can J Ophthalmol 2000;18(6):274-7.
Galin MA, Hung PT, Obstbaum SA. Cataract extraction in glaucoma. Am J Ophthalmol 2000;87(2):124-9.
Galin MA, Obstbaum SA, Asano Y et al. Laser trabeculoplasty and cataract surgery. Trans. Ophthalmol. Soc. U.K. 1984;104:72-5.
Hopkins JJ, Apel A, Trope GE et al. Early intraocular pressure after phacoemulsification combined with trabeculectomy. Ophthalmic Surg Lasers 2000;29(4):273-9.
Jacobi PC, Dietlein TS, Krieglstein GK. Comparative study of trabecular aspiration vs trabeculectomy in glaucoma triple procedure to treat pseudoexfoliation glaucoma. Arch Ophthalmol 2000;117(10):1311-8.
Luntz MH, Freedman J. The fornix-based conjunctival flap in glaucoma filtration surgery. Ophthalmic Surg 2000;11(8):516-21.
Maltzman BA, Agin M. Argon peripheral iridotomy and cataract formation. Ann Ophthalmol 2000;20(1):28-30.
O'Connell EJ, Karseras AG. Intraocular surgery in advanced glaucoma. Br J Ophthalmol 2000;60(2):124-31.
Park HJ, Kwon YH, Weitzman M et al. Temporal corneal phacoemulsification in patients with filtered glaucoma. Arch Ophthalmol 2000;115(11):1375-80.
Raitta C, Vesti E. The effect of sodium hyaluronate on the outcome of trabeculectomy. Ophthalmic Surg 2000;22(3):145-9.
Shammas HJ. Anterior intraocular lens dislocation after combined cataract extraction trabeculectomy. J Cataract Refract Surg 2000;22(3):358-61.
Sharpe ED, Simmons RJ. A prospective comparison of Amvisc (TM) and Healon (Reg.trademark) in cataract surgery. J Cataract Refract Surg 2000;12(1):47-9.
Shin DH, Hughes BA, Song MS et al. Primary glaucoma triple procedure with or without adjunctive mitomycin. Prognostic factors for filtration failure. Ophthalmology 2000;103(11):1925-33.
Shin DH, Simone PA, Song MS et al. Adjunctive subconjunctival mitomycin C in glaucoma triple procedure. Ophthalmology 2000;102(10):1550-8.
Stewart RH, Loftis MD. Combined cataract extraction and thermal sclerostomy versus combined cataract extraction and trabeculectomy. Ophthalmic Surg 2000;7(3):93-5.
Tezel G, Kolker AE, Kass MA et al. Late removal of releasable sutures after trabeculectomy or combined trabeculectomy with cataract extraction supplemented with antifibrotics. J Glaucoma 2000;7(2):75-81.
Veldman E, Greve EL. Glaucoma filtering surgery, a retrospective study of 300 operations. Doc Ophthalmol 2000;67(1-2):151-70.
Yamagami S, Araie M, Mori M et al. Posterior chamber intraocular lens implantation in filtered or nonfiltered glaucoma eyes. Jpn J Ophthalmol 2000;38(1):71-9.
| Reason for Exclusion | Number of Articles |
|---|---|
| Does not report outcomes relevant to our study questions. | 6 |
| Was neither a controlled trial nor a case series ≥ 100 patients. | 4 |
| No original data. | 3 |
| Only addresses intracapsular cataract extraction. | 3 |
| Data reported in an unabstractable fashion. | 2 |
| Other (e.g., more than half the patients did not have glaucoma) | 2 |
| Does not address open angle or primary angle closure glaucoma. | 1 |
| Study Question Addressed | Number of Articles a |
|---|---|
| Effect of cataract surgery on long-term IOP control in open angle glaucoma patients. | 43 |
| Effect of cataract surgery on short-term IOP control in open angle glaucoma patients. | 15 |
| Effect of glaucoma surgery on the development and progression of cataract. | 22 |
| Risks and benefits of staged versus combined procedures. | 5 |
| Risks and benefits of anti-fibrosis agents. | 16 |
| Risks and benefits of single-site versus two-site operations. | 7 |
| Risks and benefits of trabeculectomy versus endoscopic laser versus deep sclerectomy/viscocanulostomy. | 3 |
| Risks and benefits of nuclear expression versus phacoemulsification for patients with cataract and glaucoma. | 11 |
An article may address more than one topic.
| Topic | # of Studies | # of Eyes Enrolled Range (mean/median) | Mean Age Range (mean/median) | Proportion of Patients African-American Range (mean/median) | Mean Reported Followup Range (median) | Quality Score Mean (Range) |
|---|---|---|---|---|---|---|
| Long-Term IOP Control | 44 | 19 to 310 (110/88) | 63 to 81 (74/74) | 0.0 to 0.17 (0.06/0.03) | 6–11 months to ≥24 months (12–23 months) | 50%a (2% to 83%) |
| Short-Term IOP Control | 15 | 16 to 310 (96/68) | 69 to 80 (74/75) | 0.07 to 0.31 (0.14/0.11) | ≤24 hours to 12–23 months (6–11 months) | 55%b (23% to 75%) |
| Development of Cataract after Glaucoma Surgery | 22 | 22 to 776 (175/140) | 50 to 72 (62/63) | 0.05 to 1.0 (0.29/0.12) | 3–5 months to ≥24 months (12–23 months) | 64%c (23% to 88%) |
| Risks/Benefits of Anti-Fibrosis Agents | 16 | 24 to 174 (62/44) | 50 to 76 (71/74) | 0.03 to 0.77 (0.36/0.03) | 3–5 months to ≥24 months (12–23 months) | 61%d (33% to 80%) |
| One-Site versus Two-Site Procedures | 7 | 20 to 132 (68/50) | 65 to 78 (74/75) | Reported for only one study. | 6–11 months to 12–23 months (1–-23 months) | 40%e (10% to 64%) |
| Nuclear Expression versus Phacoemulsification | 11 | 47 to 397 (117/70) | 71 to 78 (76/76) | 0 to 0.11 (0.04/0) | 6–11 months to ≥24 months (12–23 months) | 40%f (13% to 54%) |
| Staged versus Combined Procedures | 5 | 40 to 106 (76/79) | 50 to 76 (67/71) | 0.08 to 0.24 (0.16/0.16) | 12–23 months to ≥24 months (12–23 months) | Calculated only for one study. |
| Trabeculectomy versus Other Glaucoma Procedures | 3 | 58 to 70 (63/60) | 78 to 80 (79/79) | Reported for only one study. | Reported for only one study. | 43%g (21% to 59%) |
30 studies
15 studies
13 studies
14 studies
7 studies
11 studies
3 studies
| Category | Range | Mean |
|---|---|---|
| Representativeness | 0%–94% | 58% |
| Bias and Confounding | 0%–67% | 28% |
| Therapy | 0%–100% | 65% |
| Outcomes | 0%–83% | 53% |
| Statistical Analysis | 0%–100% | 61% |
| Overall | 3%–88% | 53% |
Scores were computed only for randomized controlled trials, non-randomized controlled trials, and cohort studies.
Free Full text in PMC]Free Full text in PMC]Free Full text in PMC]Free Full text in PMC]Free Full text in PMC]
[PubMed]Free Full text in PMC]
[PubMed]Free Full text in PMC]Free Full text in PMC]Free Full text in PMC]Free Full text in PMC]
[PubMed]Free Full text in PMC]Free Full text in PMC]Free Full text in PMC]Free Full text in PMC]
[PubMed]Free Full text in PMC]Robinson KA, Hinegardner PG, Lansing P. Development of an optimal search strategy for the retrieval of controlled trials using PubMed [abstract]. Proceedings of the 6th International Cochrane Colloquium; 1998 Oct 22–26; Abstr Workshops Sci Sess Int Cochrane Coll. 1998;6:85.
