Efficacy of bilateral lateral rectus recession versus unilateral recession and resection for basic-type intermittent exotropia: a meta-analysis

Authors: Wang X, Zhu Q, Liu L.

Geographical coverage: No geographical limits to search. Of 10 included studies, three were conducted in China, two in Korea and one in Brazil.

Sector: Biomedical

Sub-sector: Treatment

Equity focus: None

Study population: General population

Review type: Effectiveness review

Quantitative synthesis method: Meta-analysis

Qualitative synthesis method: Not applicable

Background: Bilateral lateral rectus recession (BLR) and unilateral lateral rectus recession combined with medial rectus resection in the same eye (RR) are the two most common approaches for treating patients with basic-type intermittent expropia (IXT). Many studies have investigated the effectiveness of these two approaches over past decades; however, their findings have been inconclusive.

Objectives: To investigate the efficacy of bilateral lateral rectus recession (BLR) versus unilateral recession and resection (RR) for the treatment of patients with basic-type intermittent exotropia (IXT).

Main findings:

This meta-analysis provides evidence that both the BLR and RR procedures have similar efficacy for the treatment of basic-type IXT.

From 1,243 screened articles, a total of 10 studies involving 967 patients were included in the analysis. This consisted of four RCTs and six retrospective studies.

The authors found no significant differences in success rates between the BLR and RR groups at various postoperative stages (1-day to 1-week, 6-month, and last follow-up). The rates of overcorrection and undercorrection were comparable between the two groups. Similarly, no significant differences were observed in postoperative deviation at the aforementioned stages. Sensitivity analysis indicated that the exclusion of individual studies did not significantly affect the results.

Authors note that further statistical analyses (for example, restriction to subgroups based on variable definitions of surgical success or durations of long-term follow-up) are needed to address potential bias when more studies are available.

Methodology:

Studies were included if they met the following criteria: (1) the study was a randomised controlled study or retrospective observational study; (2) the study included subjects in the studies with basic-type IXT; and (3) the study provided success rate and/or postoperative deviation for both BLR and RR study arms. Authors excluded (1) non-published articles (e.g. conference abstracts); (2) case reports or case series with no control group; (3) literature reviews, letters and correspondence; (4) studies that included patients with the A or V pattern, vertical deviation, oblique muscle dysfunction, or a history of strabismic surgeries; and (5) studies that included patients with IXT for whom the surgical outcomes for the basic type could not be extracted.

Authors searched the EMBASE, PubMed, Web of Science and Cochrane Library databases for relevant studies published before April 2020 with no language restrictions. A range of key search terms were used. All potentially eligible studies in any language were considered for review. Reference lists from reports on clinical trials, meta-analyses and systematic reviews were examined to identify relevant studies.

Two authors independently reviewed the titles and abstracts to identify potentially eligible articles, and then screened the full-text articles to determine the final inclusions. Discrepancies in inclusion were adjudicated through consultation with a third author.

Risk of bias of RCTs was assessed by using the Revised Cochrane risk of bias tool in randomised trials. Five domains of quality assessment were performed for each trial: randomisation process, deviations from intended interventions, missing outcome data, measurement of the outcome and selection of the reported result. For observational studies, the Newcastle-Ottawa Scale was applied with an 8-item scale to evaluate quality based on three criteria: Two reviewers independently assessed the risk of bias. Discussion involved a third author when discrepancies could not be resolved.

Data analysis was performed using Review Manager 5.3. The weighted mean difference (MD) with the 95% confidence interval (CI) was assessed for continuous variables. The odds ratio (OR) with the 95% CI was calculated for dichotomous variables. A random-effects model was applied in all cases. The inverse variance method was used for continuous data, and the Mantel-Haenszel approach was used for dichotomous outcomes. Statistical heterogeneity was quantified by a chi square statistic; for this test, only a result with a p-value <0.05 was considered heterogeneous and was reported in the text. Additionally, the I² statistic was used to assess the magnitude of the heterogeneity among studies. Authors also undertook a sensitivity analysis, removing one study at a time.

Applicability/external validity: Authors note that findings on the most effective approach over the past decade have been inconclusive, although they argue their review is methodologically stronger.

Geographic focus: Authors do not explore how results might differ based on geographic area; although around half of the reviews included are from low- and middle-income countries.

Summary of quality assessment:

There were some limitations to the methods used to identify, critically include, and appraise the studies. Unpublished material was not considered for inclusion and it is unclear whether relevant experts were consulted. No table was included indicating how each of the included studies had performed against the agreed quality criteria. While the methods employed to analyse the data were generally robust, it was unclear if and how the quality assessment of each study was accounted for in the analysis. For these reasons, we have medium confidence in the results of this review.