2.1. Study registration
INPLASY systematical review protocol registration number is Inplasy protocol 202080026 on 08 August 2020 .https://inplasy.com/inplasy-2020-8-0026/).
This protocol will base on the preferred reporting items for systematic reviews and meta-analyses protocols (PRISMA-P) statement guidelines.[13]
2.2. Inclusion criteria for study selection
2.2.1. Types of studies.
All the RCTs of plum-blossom needle as the treatment for myopia will be involved in the review and no language or publication status limitation will be imposed. Non-rcts, review studies, case reports and animal experiments will be excluded
2.2.2. Types of participants.
Adolescents 18 years of age or younger who meet the diagnostic criteria for myopia，and the included patients are simple myopia with a diopter of -3.00 D and below, with corrected visual acuity ≥1.0. Exclusion criteria：1. Those with other eye diseases 2. Children with mentally handicapped and mental disorders 3. Those who are participating in Chinese and Western medicine or other treatment methods such as helium-neon laser therapy, visual physiological stimulation (CAM), massage, ear acupoints, etc.
2.2.3. Types of interventions.
Research using plum-blossom needle or combined with other therapies, without limiting the treatment time and dose.
2.2.4. Types of outcome measures.
Our primary outcomes will regard myopia progression and axial elongation as efficacy criteria. Myopia progression will be assessed as mean change in refractive error, measured in diopter, per year. Mean change in axial length, measured in millimeters, per year, will also be evaluated.
2.2.4.1. Primary outcomes.
- Visual acuity score before and after treatment
- Diopter
- Eye axis length
2.2.4.2. Secondary outcomes.
- Corneal curvature
- Lens thickness
- Ciliary body thickness
2.3. Search methods for the identification of studies.
2.3.1. Electronic searches.
Using the three keywords of "myopia, acupuncture, and plum blossom needle", from the establishment of the database to July 2020, the following electronic databases will be searched regardless of their status and language: Medline, EMBASE, Web of Science, the Cochrane Library, PubMed, China National Knowledge Infrastructure (CNKI), China Biology Medicine (CBM), Chinese Scientific Journal Database (VIP), Wanfang Database, Chinese Biomedical Literature Database (CBLD), Chinese Science and Technology Periodical Database (CSTPD). RCT registration websites, including http://www.ClinicalTrials.gov and http://www.chictr.org.cn, will also be searched. Review Manager V.5.4 will be used for statistical analysis. Two reviewers will independently select studies, extract and code the data, assess risk of bias of the included studies, and evaluate the quality of evidence for outcomes.
2.3.2. Other searches.
Taking account of possible omission, not only the published studies in journals but also gray literatures will be retrieved, mainly through conference papers and references.
2.4. Data collection and analysis
2.4.1. Selection of studies.
We plan to conduct this systematic review between 30 October 2020 and 30 July 2022. All reviewers have undergone a training to ensure a basic understanding of the background and purpose of the review. The selection of studies will be accomplished by two researches independently and be cross-checked. For literature management, EndnoteX9 will be utilized, with which the collected literatures will be imported and the duplicate ones will be deleted. Firstly, researchers will screen the titles and abstracts of the articles and exclude the distinctly ineligible ones. Afterward, the full text will be attentively inspected according to the inclusion and exclusion criteria previously set up. A third researcher will take part in the discussion and make the arbitration where there are discrepancies. The whole selection process will be presented in a PRISMA flow diagram (Fig. 1).
2.4.2. Data extraction and management.
Using a predefined data collection form, two researchers will complete the data
extraction. Extracted data shall include but not be limited to the following items: title, first author, publication time, sample size, severity and duration of myopia, age and gender of participants, outcomes, and adverse events. Inconsistencies will be resolved through discussion and consultation with a third researcher.
2.4.3. Coping with the questionable data.
We will try to contact the first or corresponding authors of the included studies by telephone or email to retrieve missing or insufficient trial data. If missing data are unavailable, we will make an assumption using the terms “missing at random” and
“not missing at random” to represent different scenarios. For the data “missing at random”, only the available data will be analyzed. For the data “not missing at random”, we will displace the missing data with replacement values and a sensitivity analysis will be used to determine whether the results are inconsistent.
2.4.4. Assessment of risk of bias in included studies.
Two reviewers will separately appraise the risk of bias of the involved studies according to Cochrane Handbook for Systematic Reviews of Intervention. There will be three levels of evaluation results utilizing three scores of “L,’, “U,” and “H,” respectively, indicating low-risk, uncertain, and high-risk. In the appraisal, seven sectors will be assessed, including random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, completeness of outcome data, selective reporting, and other sources of bias. Discussion with a third reviewer will be conducted to cope with potential divergence, and the corresponding author will be connected when needed.
2.4.5. Measures of treatment effect.
To appraise the treatment effect for continuous data, mean difference (MD) in terms of 95% confidence interval (CI) will be applied, just as relative risk (RR) for dichotomous data.
2.4.6. Assessment of heterogeneity.
On the basis of the data analysis, random effect or fixed effect models will be employed according to the heterogeneity given by I2 statistic value. To be concrete, a fixed effect model will be adopted if the heterogeneity is indicated as high(I2<50%); otherwise, a random effect model will be applied on the contrary.
2.4.7. Assessment of reporting bias.
The reporting bias will be visually indicated by funnel plots when studies are sufficient (at least 10 RCTs). If asymmetry is shown, Begg’s and Egger’s test will be completed and value of P>0.05 will be interpreted as no substantial reporting bias. Since the asymmetry of the funnel plot is not a substitute to publication bias, the potential reasons for it will be discriminated with terms like small sample size, low
methodological quality, or true heterogeneity.
2.4.8. Data synthesis.
The systematic review will be conducted with the use of RevMan 5.4. Taking account of the heterogeneity assessment, MD or RR with fixed or random effect model will be computed. Additionally, if heterogeneity is considered significant, the sensitivity or subgroup analysis will be generated to distinguish the source of it. When it comes to the situation that the data are insufficient for quantitative analysis, the review will only represent and summarize the evidence.
2.4.9. Sensitivity analysis.
As is mentioned above, sensitivity analysis will be done when the heterogeneity is greater than 50%. Concretely, the meta-analysis will be reconducted after the studies with low quality and small sample size are excluded to identify whether these factors influence the result.
2.4.10. Subgroup analysis.
Subgroup analysis will also be performed after substantial heterogeneity is observed to find out the reasons. Characteristics like outcome type, disease duration, study quality, race, and so forth will be the content of the subgroup analysis.
2.4.11. Quality of evidence.
The Grading of Recommendations Assessment, Development and Evaluation (GRADE) working group methodology will be applied for the quality of evidence for all outcomes. [14] Six domains will be assessed, containing risk of bias, consistency,
directness, precision, publication bias and additional points. The assessments will be categorized into 4 levels: high, moderate, low, or very low.