Efficacy of atropine, orthokeratology, and combined atropine with orthokeratology for childhood myopia: A systematic review and network meta-analysis

Authors: Tsai HR, Wang JH, Huang HK, Chen TL, Chen PW, Chiu CJ.

Geographical coverage: Asia

Sector: Biomedical

Sub-sector: Treatment

Equity focus: Children aged <18 years.

Study population: Children with myopia.

Review type: Effectiveness review

Quantitative synthesis method: Network meta-analysis

Qualitative synthesis method: Not applicable

Background: Orthokeratology (Ortho-K), atropine eye drops and combined atropine with Ortho-K are proven to be effective ways to prevent myopic progression in many studies, but there is scarce evidence regarding the comparative efficacy of different dosages of atropine, Ortho-K, and combined atropine with Ortho-K for childhood myopia.

Objectives: To evaluate the comparative efficacy in myopia control among different atropine dosages, Ortho-K, and combined use of atropine with Ortho-K – to provide information to help facilitate decision-making in clinical practice.

Main findings:

Overall findings of this review showed that atropine (0.01%-1%), Ortho-K and 0.01% atropine combined with Ortho-K could significantly slow down myopia progression. There is a dose-related pattern in the atropine efficacy for myopia control.

A total of 19 randomised controlled trials (3,435 patients) were included in the review. Among the 19 trials, two studies had four arms, two studies had three arms, and the remaining studies were two-armed. All studies were conducted in the Asian region. Regarding the follow-up period, six RCTs were followed-up for two years while the remaining RCTs were followed up for one year.

In terms of studies’ quality, there were several concerns across a number of studies, including lack of descriptions about the randomisation methods and blinding of the participants or outcome assessors. Regarding missing outcome data, several articles reported a high drop-out rate; therefore, they were considered to have a high risk of bias. Generally, most studies were rated as having some concern in the domain of overall bias.

Network meta-analysis revealed that 0.01%-1% atropine, Ortho-K and 0.01% atropine combined with Ortho-K inhibited axial elongation (AL) over one year. For refractive change, SUCRA analysis revealed that the hierarchy was high-dose (0.5%-1%), moderate-dose (0.1%-0.25%), and low-dose (0.01%-0.05%) atropine. Regarding AL, SUCRA analysis revealed the following hierarchy: Ortho-K combined with 0.01% atropine, high-dose atropine, moderate-dose atropine, Ortho-K, and low-dose atropine.

Authors found no significant funnel plot asymmetry, which indicated the absence of small-study effects in the network. Furthermore, the Egger test revealed no significant publication bias in AL (p Z 0.91) but in SER (p <0.1).

The information provided in the review could provide helpful guidance regarding the management of myopic progression which may be relevant for policy.

Methodology:

The inclusion criteria were as follows: (1) RCTs; (2) treatment modalities including placebo, atropine eye drops, Ortho-K, or combined use of Ortho-K and atropine eyedrops for slowing myopic progression; (3) participants with a myopia diagnosis who were aged <18 years (4) mean follow-up period S one year; (5) outcomes of interest including mean annual change in standardised equivalent refractive error (SER) (dioptre/year) or AL (millimetres/ year). Studies were excluded if (1) they were review articles, case reports, case series, animal or laboratory studies, or conference abstracts; and (2) if they lacked the required outcome measures.

A literature search was conducted to identified RCTs describing the efficacy of different atropine doses and Ortho-K in myopia control published before August 2021 in the PubMed, EMBASE (Ovid), and Cochrane Library databases. There were no language restrictions. The keywords “atropine”, “myopia control” and “orthokeratology”, “combined atropine with orthokeratology”, as well as their synonyms and derivatives, were used. Two independent authors reviewed all abstracts, studies and retrieved citations. The reference sections of the retrieved articles were assessed to identify more relevant studies.

Two authors independently screened the titles and abstracts; moreover, they assessed the remaining full articles based on the eligibility criteria. Given that the most objective evidence for evaluating the effect of Ortho-K is the decrease in the rate of globe AL, authors mainly assessed treatments involving Ortho-K based on its AL inhibition effect. The control group was comprised of patients who received single vision spectacle lens and tropicamide eyedrops. In cases where specific aspects required clarification, efforts were made to contact the corresponding authors for further information. For studies that did not report the standard deviation, authors calculated the standard deviation using formulas described in the Cochrane Handbook for systematic reviews of interventions.

The methodological quality of RCTs was evaluated using the Cochrane Collaboration’s risk-of-bias assessment tool by two reviewers, with disagreements being resolved through discussions with two other reviewers.

Statistical analyses were performed using Stata and R software. The effect size was presented as the mean difference (MD) with 95% confidence intervals (CIs) for continuous outcomes (SER and AL). Network Meta-Analysis (NMA) was conducted using frequentist random-effects models and a contrast-based approach. Global and local inconsistencies were evaluated. The R package netmeta was used for visualisation. SUCRA curves ranked treatments, with scores ranging from 1 (best) to 0 (worst). Sensitivity analyses tested the robustness of findings by excluding high-bias studies. Egger’s tests examined potential publication biases, with p-value <0.1 indicating bias. Subgroup analysis stratified atropine dosage into high, moderate, and low doses for further NMA.

Applicability/external validity: Authors note that their findings reflect those of previous work, while emphasising that much of their detailed focus (to comparing treatment dosages) is new.

Geographic focus: All included studies were published in Asia (no further geographic information provided).

Summary of quality assessment:

While many of the approaches to identifying and critically appraising studies were robust, with two authors undertaking all key asks, there was no attempt to include unpublished material, no tabular breakdown was provided of the quality assessments of individual studies and the start date of the search was unclear. There was no discussion of the heterogeneity of included studies or consideration of unit of analysis errors, which we feel should have been considered due to the inclusion of RCTs. In addition, we note that concerns regarding bias were associated with many of the included studies and that the findings are not generalisable beyond Asia, due to a lack of diversity in the regions where the studies were conducted. For these reasons, we have low confidence in the findings of this review.