The role of feedback in amblyopia treatment – a multi-centre randomised control trial

BACKGROUND: Conventional occlusion is an effective treatment for amblyopia; however, adherence remains a significant barrier. This is the first randomised controlled trial (RCT) investigating whether objective, electronically monitored adherence "feedback" improves amblyopia treatment adherence. METHODS: This unmasked, parallel two-armed multicentre RCT included 102 children aged 3-8 yrs with monocular amblyopia (≥0.3 logMAR interocular difference). Participants could have up to 18 months of previous patching and were prescribed 10 h/6days of patching. Participants were randomised to a Feedback Group (n = 51), receiving feedback from treatment monitors, or Controls (n = 51). Change in adherence (CA) was measured from the first to last available monitor for patching and glasses over 12 weeks. Regression models explored factors influencing visual outcome and adherence. RESULTS: Of 102 participants, 74 were analysed for patching and 78 for glasses. Mean patching CA was -0.39 ± 2.01 h/day (control) versus -0.32 ± 2.20 h/day (feedback), with no significant group difference (P = 0.89). Median glasses-wearing CA was -0.55 IQR:2.55 h/day (control) vs. -0.05 IQR:1.73 h/day (feedback), also non-significant (P = 0.38). Overall average adherence to glasses was 10.3 h/day and 7.9 h/day for patching. Younger age, less previous patching, and higher adherence to treatment significantly predicted better visual outcome. Females had significantly lower glasses-wearing adherence. CONCLUSION: This study shows for the first time that patching and glasses adherence can be monitored and fed back to patients and their carers. While we found no additional influence of feedback on adherence, we observed that when children and their guardians were aware of active monitoring and frequently seen, we observed high and sustained levels of adherence. The significant correlations to visual outcomes further highlight the importance of early treatment in amblyopia.