An updated meta-analysis of controlling myopia with auricular acupoint stimulation

Author: Sangvatanakul P, Tangthianchaichana J, Tasanarong A, Pabalan N, and Tharabenjasin P.

Geographical coverage: China

Sector: Biomedical

Sub-sector: Treatment

Equity focus: Teenagers

Study population: People with myopia treated with auricular acupoint stimulation or another technique (controls)

Review type: Effectiveness review

Quantitative synthesis method: Meta-analysis

Qualitative synthesis method: Not applicable

Background: Myopia, particularly prevalent among young people in Asia, is being addressed through various methods, including auricular acupoint stimulation (AAS)1. This technique, a part of Chinese medicine known as auriculotherapy, has been widely used for the prevention, reduction, and treatment of myopia. Past research has demonstrated significant improvements in visual acuity in patients with low-grade myopia following AAS treatment. Further studies on this topic have been published in both Chinese and English languages.

Objectives: To investigate the primary data regarding the effect of AAS intervention on myopia, using a meta-analysis, to clarify the effectiveness of AAS for preventing or retarding the progression of myopia.

Main findings:

In summary, the meta-analysis provided evidence supporting the effectiveness of AAS in managing myopia. The substantial effect size (up to a 3.4-fold increase), robustness, and absence of bias further reinforced this finding.

An initial search of 16,459 citations was narrowed down to 12 articles, mostly in Chinese, for inclusion in the study. All subjects were Chinese, and interventions involved auricular acupuncture with variations. Control treatments primarily used eye drops. Visual acuity was the consistent outcome measure across all studies. Treatment frequencies and durations varied. Authors found low risk of bias in the random-sequence generation with conspicuity of unclear judgements for the allocation concealments and blinding criteria.

The AAS pooled effect significantly (Pa = 0.003) favoured treatment over the controls (OR: 2.87; 95% CI: 1.44-5.75). This pre-outlier outcome was heterogeneous (Pb <10–5; I² = 87%). The Galbraith plot identified two studies as the sources of heterogeneity, located above the +2 and below the -2 confidence limits. The post-outlier outcome (OR: 3.42; 95% CI: 2.58-4.53) showed reduced heterogeneity (Pb = 0.37; I² = 8%). AAS pooled effects between pre-outlier and post-outlier generated the following outcomes: (1) intensified significance (Pa <10-5) and (2) increase in magnitude (from 2.8-fold to 3.4-fold). The two significant (Pa <0.05) AAS outcomes were robust, with none of the component studies influencing the pooled ORs. The Egger’s regression test showed no evidence of publication bias (Intercept: -0.45; p = 0.84).

Methodology:

The inclusion criteria were as follows: (1) they had to be case-control studies that examined the relationship between AAS and myopia; and (2) they needed to provide enough data to calculate the odds ratios (ORs) and 95% confidence intervals (CIs). Studies were excluded if they: (1) did not focus on myopia; (2) were reviews; or (3) lacked control groups or had unusable or missing data.

Authors searched Medline, PubMed, Google Scholar, China National Knowledge Infrastructure and Wanfang databases for association studies as of 15 May 2020. The terms used were “acupuncture ear”, “auricular acupuncture”, “auricular acupressure”, “auricular acupoint stimulation”, and “auricular myopia” as medical subject headings and texts, unrestricted by language. References cited in the retrieved articles were also screened manually to identify additional eligible studies. In cases of duplicate articles, the ones with later publication dates were selected. Primary study authors were contacted to obtain more information on incomplete data. Chinese-language articles were translated to English by the first and second authors.

Two investigators independently extracted data; disagreements were adjudicated by a third investigator, resulting in a consensus.

Risk of bias was assessed based on the Cochrane Handbook guideline, which was applied to randomised controlled trials and studies that involved interventions. Judgements were assigned by two of the authors working independently, and discrepancies were remedied through discussions with a third investigator to obtain a consensus.

The authors hypothesized that Auricular Acupoint Stimulation (AAS) reduces myopia risk and slows its progression. Odds Ratios (ORs) were calculated for each study, with ORs >1.0 favouring AAS. Heterogeneity between studies was assessed using a chi-square-based Q-test and the I² statistic. A fixed-effects model was used if studies showed functional similarities; otherwise, a random-effects model was applied1. Outliers were identified and excluded for reanalysis. Sensitivity analysis was performed by omitting one study at a time. Publication bias was assessed using Egger’s regression test.

Applicability/external validity:

The authors do not explicitly report the applicability and external validity of their findings. The authors compared their findings with a previous meta-analysis conducted in Chinese. The current study, which included English-language articles and had a larger sample size, found a moderated effect of 2.8 to 3.4-fold in contrast to the 4.1-fold effect found in the previous study. The current analysis also assessed the risk of bias and treated outliers, steps not taken in the previous study.

Geographic focus:

The search was not geographically limited and given that the treatment of interest is rooted in Chinese medicine, it’s unsurprising that all subjects in the study are Chinese. However, this implies that the results may not be universally applicable.

Summary of quality assessment:

The methods used to select, include, and critically evaluate studies had several shortcomings. There’s no indication that unpublished materials were considered for inclusion, and it’s unclear if the search was constrained by time. The review doesn’t specify if multiple reviewers were involved in deciding which studies to include. Although the data analysis was generally robust, the authors didn’t explain how they addressed potential unit of analysis errors, which could have been relevant for some of the included RCTs. Due to these issues, our confidence in the review’s findings is low.