Efficacy And Safety Of Behavioral Training For Amblyopia:a Systematic Review And Meta-analysis

Objective:Based on systematic review and meta-analysis of randomized controlled trials,evaluating the efficacy and safety of behavior training for amblyopia systematically.Meanwhile,it provides ideas for follow-up research and the design of behavior training.Methods:Electronic databases(PubMed,Embase,Web of Science,EBSCO,Cochrane Library,CBM,CNKI,WANFANG DATA,CQVIP,WHO ICTRP)were systematically searched from inception to August 2021.Visual acuity or best corrected visual acuity as well as safety were considered the primary outcomes in this analysis.The secondary outcomes were stereopsis,visual evoked potential amplitude and visual evoked potential latency.Results:A total of 25 randomized controlled trials involving 2232 participants were included.Among them 10 trials were perceptual learning training,3 trials were the combined treatment of perceptual learning training and occlusion therapy,9 trials were video game training,and 3 trials were the combined treatment of video game training and occlusion therapy.The overall quality of the included studies ranged from moderate to high.Effectiveness of behavior training in the treatment of amblyopia:in terms of visual acuity or best corrected visual acuity,the pooled results of ten trials suggested that the difference between perceptual learning training and control group was statistically significant(SMD:2.02;p=0.00;95%CI:1.44-2.59).Three trials suggested that there was no significant difference between the combined treatment of perceptual learning training and occlusion therapy(SMD:0.94;p=0.00;95%CI:-0.08-1.95).Nine trials suggested that the difference between video game training and control group was not statistically significant(SMD:0.04;p=0.02;95%CI:-0.18-0.26),the same as three trials of the combined treatment of video game training and occlusion therapy(SMD:0.45;p=0.72;95%CI:0.14-0.76).In terms of stereopsis,the pooled results of one trial suggested that the difference between the combined treatment of perceptual learning training and occlusion therapy and control group was statistically significant(SMD:0.64;95%CI:0.36-0.92).Two trials suggested that the difference between video game training group and control group was not statistically significant(SMD:0.28;p=0.65;95%CI:-0.29-0.33).Besides,two trials of combined treatment of behavior training and occlusion therapy reported the improvement of stereopsis.In terms of visual evoked potential amplitude and visual evoked potential latency,the pooled results of six trials suggested that the difference between perceptual learning training and control group in improving visual evoked potential amplitude was statistically significant(SMD:1.00;p=0.02;95%CI:0.69-1.31).Six trials suggested that the difference between perceptual learning training and control group in improving visual evoked potential latency was statistically significant(SMD:-1.72;p=0.00;95%CI:-2.90--0.54).Safety of behavior training in the treatment of amblyopia:seven trials reported adverse reaction events(5 trials were video game training,1 trial was the combined treatment of video game training and occlusion therapy,and 1 trial was perceptual learning training).Other trials did not mention any information about adverse reactions.Adverse reaction events mainly fall into two categories.Firstly,it is relatively serious eye adverse reactions,including diplopia and new heterotopia.Secondly,minor adverse eye reactions or discomfort caused by auxiliary tools,including transient asthenopia,virtual reality headset intolerance,and skin irritation caused by occlusion therapy in combined treatment.Conclusion:1.The overall quality of the included trials ranged from moderate to high,some of them need to report the allocation of hidden schemes,potential other bias risks and other contents in more detail.2.Behavior training is relatively safe in the treatment of amblyopia.3.Perceptual learning training can improve the visual acuity or the best corrected visual acuity,visual evoked potential amplitude and visual evoked potential latency of amblyopia patients,while video game training has no obvious evidence for the time being.Due to the high heterogeneity of the trials and the inconsistency of the results,behavioral training is not adequate to replace sufficient traditional therapy,but it seems reasonable to be considered as an auxiliary supplementary training in clinical practice.In the future,more high-quality and large-scale RCTs are needed to investigate the efficacy and safety.