Under-correction or full correction of myopia? A meta-analysis

Authors: Yazdani N, Sadeghi R, Ehsaei A, Taghipour A, Hasanzadeh S, Zarifmahmoudi L, Heravian SJ.

Geographical coverage: Not reported

Sector: Biomedical

Sub-sector: Treatment

Equity focus:  None stated

Study population: People with myopia aged between 6-33 years.  

Review type: Effectiveness review

Quantitative synthesis method: Meta-analysis

Qualitative synthesis method: Not applicable

Background: Under-correction is one potential intervention to reduce or slow myopia progression, with one hypothesis being that under-correction may decrease the accommodative effort and the lag of accommodation and subsequently slow or reduce myopia progression. However, studies examining this hypothesis to date have shown contradictory findings.

Objectives: To compare the effect of full correction versus under-correction on myopia progression.

Main findings

Overall findings of this review show that myopic eyes which are fully corrected with non-cycloplegic refraction with maximum plus sphere, are less prone to myopia progression, in comparison to those which were under corrected (although this finding is statistically, but not clinically, significant).

The meta-analysis conducted by the authors encompassed six studies, including two randomized controlled trials (RCTs) and four non-RCTs, with a total of 695 participants aged between 6 and 33 years. These participants were divided into two groups: a full-correction group (n = 371) and an under-correction group (n = 324).

When cycloplegic refraction was used, authors reported that the average myopia progression was slightly higher in the full correction group compared to the under correction group, but the difference was not statistically significant (p = 0.085).

In studies using non-cycloplegic subjective refraction, authors found that the under-correction group showed a slightly higher myopia progression compared to the full-correction group, but again, the difference was not statistically significant (p = 0.175).

In conclusion, neither full correction nor under correction significantly affected the progression of myopia, regardless of the type of refraction used.

Authors identify various limitations: some of the included studies use only one specific level of under-correction, making comparison across the board difficult. The results showed high heterogeneity which is likely to arise through differences in inclusion criteria, amount of under-correction and types of refraction, while study quality might affect the heterogeneity of the results. Authors conclude that, “any conclusions are limited by the characteristics of the studies”.

Authors indicate that, regarding cycloplegic refraction, further studies are needed to better understand these trends.

Methodology:

Authors performed a literature search in PubMed, Scopus, Science Direct, Ovid, Web of Science and Cochrane Library. No date limit or language restriction was used; the search was updated until July 2019. Used a search algorithm that was based on a combination of the terms: ‘‘Under-correction”, “Full correction”, “Myopia” and “Nearsightedness’’. The Boolean operators ‘‘AND’’ and ‘‘OR’’ were used to specify the search. References of the retrieved articles were also screened by two independent authors. Two researchers independently reviewed identified studies against the inclusion/exclusion criteria.

For each study, information was extracted regarding basic study data (authors, journals, year of publication, country of origin, and study design), patient characteristics (mean age, gender, number of patients) and technical aspects (amount of under-correction, type of refraction and applied device). Methodological quality assessment of the literature was evaluated according to the Critical Appraisal Skills Program.

Statistical analysis was performed using Comprehensive Meta-Analysis (version 2, Biostat Inc., USA). For each study, the mean difference in myopia progression recorded in dioptre notation was determined for the full correction and under-correction groups. To pool the effect sizes across studies, a random-effects model was used.

Heterogeneity was evaluated by the Cochrane Q test (the significance level was considered to be 0.05.), and I² index. Publication bias was evaluated graphically by funnel plots and statistically by Egger’s regression intercept method.

Applicability/external validity: The limitations identified by the authors, specifically in relation to the methodologies used for measurements and the quality of the studies included, suggest that the conclusions drawn may have restricted relevance and reliability.

Geographic focus: Not reported

Summary of quality assessment:

While the search for evidence was reasonably comprehensive, there was no evidence of unpublished studies being considered for inclusion or relevant experts consulted. There was a lack of clarity regarding the inclusion criteria in terms of the types and studies, participants and settings that were considered for inclusion. It was unclear whether data was independently extracted by two reviewers and there was felt to be insufficient consideration of the differences between reported studies and how these may have contributed to the outcomes that were reported. For these reasons, we attributed low confidence in the findings of this review.