Cochrane review: Interventions for unilateral refractive amblyopia

Background Unilateral refractive amblyopia is a common cause of reduced visual acuity in childhood, but optimal treatment is not well defined. This review examined the treatment effect from spectacles and conventional occlusion. Objectives Evaluation of the evidence of the effectiveness of spectacles and or occlusion in the treatment of unilateral refractive amblyopia. Search strategy We searched the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, EMBASE and LILACS. Relevant conference proceedings were manually searched. There were no date or language restrictions. The searches were last run on 7 July 2008. Selection criteria Randomised controlled trials of treatment for unilateral refractive amblyopia by spectacles, with or without occlusion were eligible. We included studies with participants of any age. Data collection and analysis Two authors independently assessed abstracts identified by the searches. We obtained full text copies and contacted study authors where necessary. Eight trials were eligible for inclusion. Data were extracted from seven. No meta-analysis was performed. Main results For all studies mean acuity (standard deviation (SD)) in the amblyopic eye post treatment is reported. Comparison: Spectacles only versus no treatment (Clarke 2003). Mean (SD) visual acuity: spectacles group 0.31 (0.17); no treatment group 0.42 (0.19). Mean difference (MD) between groups -0.11 (borderline statistical significance: 95% confidence interval (CI) -0.22 to 0.00). Comparison: Spectacles plus occlusion versus no treatment (Clarke 2003). Mean (SD) visual acuity: full treatment 0.22 (0.13); no treatment 0.42 (0.19). Mean difference between the groups -0.20 (statistically significant: 95% CI -0.30 to -0.10). Comparison: Spectacles plus occlusion versus spectacles only: Clarke 2003 MD -0.09 (borderline statistical significance 95% CI, -0.18 to 0.00); PEDIG 2005b; MD -0.15 (not statistically significant 95% CI -0.32 to 0.02); PEDIG 2006a; MD 0.01 (not statistically significant 95% CI -0.08 to 0.10). Comparison: Occlusion regimes. PEDIG 2003a: 2 hours versus 6 hours for moderate amblyopia: MD 0.01 (not statistically significant: 95% CI -0.06 to 0.08); PEDIG 2003b: 6 hours versus full-time for severe amblyopia: MD 0.03 (not statistically significant: 95% CI -0.08 to 0.14). Stewart 2007a: 6 hours versus full-time occlusion: MD -0.12 (not statistically significant: 95% CI -0.27 to 0.03) Authors' conclusions In some cases of unilateral refractive amblyopia it appears that there is a treatment benefit from refractive correction alone. Where amblyopia persists there is some evidence that adding occlusion further improves vision. It remains unclear which treatment regimes are optimal for individual patients. The nature of any dose/response effect from occlusion still needs to be clarified. Plain Language Summary Patching treatment for amblyopia (lazy eye) caused by needing spectacles Patching treatment for amblyopia (lazy eye) caused by unequal optical errors in either eye when the (usually) worst eye can fail to develop optimum vision. Amblyopia is a condition where the vision of one (or more rarely both) eyes is reduced because the part of the brain responsible for vision does not develop properly. Amblyopia happens because the brain receives a weaker image from the eye with the greater optical error and prefers the eye with a clearer image. Children who have amblyopia because of optical errors are often given a patch to cover the good eye in addition to their spectacle correction to improve their vision. This review found that for some children with this type of amblyopia a period of spectacles wear can restore normal vision. At present it is not possible to tell at the start of treatment which children will just need spectacles and which ones will need a patch as well. For those children who still have reduced vision even after they have been wearing their spectacles for a while there is evidence that patching therapy further improves their vision. The amount of patching that will be needed for an individual child can not yet be predicted. This is because the effects of factors such as age are not fully understood. These findings are based on the results of six high quality trials.