Statistics from Altmetric.com
More evidence that we need evidence
The evidence base for amblyopia treatments is not well developed.1,2 A study by Tan et al in this issue of the BJO (p 291) documents differences in amblyopia treatments between the UK and Germany, Switzerland, and Austria (“German speaking countries”). A recent study from the United States also showed significant differences in treatments between centres.3
What gaps in our evidence base for amblyopia treatment does the study by Tan et al expose?
The upper age limit for treatment of newly diagnosed cases was significantly higher in German speaking countries than in the United Kingdom. Our knowledge of the responsiveness to treatment of each type of amblyopia in each age group is limited.4–7
The dose of part time occlusion therapy prescribed was significantly greater in German speaking countries than in the UK. There were also differences in the circumstances in which full time occlusion was prescribed. When treatment failed, occlusion treatment was discontinued at a younger age in the UK than in German speaking countries. Only limited studies of occlusion dose-response have been published.8,9
Spectacles alone, in place of spectacles plus occlusion, were used more widely in the UK than in German speaking countries for the initial treatment of severe anisometropic amblyopia. Only limited studies of spectacles only treatment of amblyopia are available in the literature at present.10 In addition, the therapeutic effects of spectacle wear may have confounded studies of occlusion treatment in the past .
Atropine penalisation was more likely to be used as a first line treatment for amblyopia in German speaking countries than in the UK. However, occlusion was used much more widely than atropine penalisation in all countries. Only limited studies of atropine penalisation have been performed,10–13 until very recently.14
Orthoptists in all countries believed a negative psychological effect of occlusion therapy was infrequent. While the negative psychological impact of amblyopia has been studied,15 very little work on the possible negative psychological effects of occlusion treatment has been undertaken.
Despite differences in treatments, orthoptists from all countries gave very similar estimates of the success rates of treatment. Actual treatment outcomes, however, remain entirely unknown in most departments. Once again these results show that current practice is based on value beliefs held by ophthalmologists and orthoptists rather than on measured evidence.
A number of very important amblyopia treatment trials are currently under way, and many of the questions raised in the study by Tan et al will be answered during the next 3 years. Recent advances in methodology have made these trials possible.16 Definitions of the types of amblyopia have been standardised.14 The use of logMAR visual acuity charts has become widespread, and validated test strategies have been developed.17 The need for robust baseline visual acuity measurements16,17 and spectacle adaptation have been recognised.16 Compliance with occlusion therapy may now be measured in a reliable way.16,18
In the future, atropine penalisation may be the first line treatment for the majority of children with amblyopia
Occlusion treatment for strabismus was first described in 1722.19 A randomised trial of treatment versus no treatment (ever) is probably not justified, as there is sufficient evidence that treatment leads to improved visual acuity in the majority of patients,5 that this improvement is maintained,20 and that the natural history is not one of spontaneous improvement.21 Clarke and collaborators in north east England have recently completed a multicentre trial that includes a no treatment arm for a limited period. Patients with anisometropic amblyopia were randomised to no treatment, spectacles alone, or spectacles plus occlusion.22 The no treatment group was crossed over to treatment after 1 year.
Moseley, Fielder, and colleagues at Imperial College London have painstakingly developed robust treatment outcome methodology over a number of years.10,16,18,23 Careful assessment of baseline logMAR visual acuity,16 a prolonged period of spectacle adaptation,10,16,23 and electronic monitoring of treatment compliance16,18,23 are the key elements of the Monitored Occlusion Treatment of Amblyopia Study (MOTAS). Pilot results have been published,23 and preliminary outcome data were presented at ARVO in May 2002.24 The dose-response relation for occlusion therapy was measured in an observational study of 4–6 year old children with strabismic, anisometropic, or combined strabismic and anisometropic amblyopia. Compliance with occlusion treatment was poor—in the order of 50%. This finding must raise doubts about the reliability of other occlusion treatment outcome trials that do not use continuous electronic compliance monitoring.
The dose-response relation for occlusion therapy was similar for all three types of amblyopia studied. Children under 6 years of age showed a more rapid rate of improvement than older children—85% of improvement occurred during the first 6 weeks of treatment. The dose-response relation was linear during the first 160 hours of treatment. Improvement plateaued after 6 months.
These results are extremely valuable and peer reviewed publications from these studies are eagerly awaited.
A series of large multicentre amblyopia treatment study (ATS) trials are now under way in North America, under the chairmanship of Jonathan Holmes.25 The studies are funded by the National Institute of Health, and undertaken by the Pediatric Eye Disease Investigator Group.26 Initial methodology studies have been completed.17,27,28 The 6 month outcome of ATS 1 was published recently.14 This was a trial of atropine penalisation versus occlusion treatment for moderate amblyopia (20/40–20/100) in 3–6 year old children. The treatments had a very similar outcome at 6 months. However, parents (slightly) preferred atropine penalisation to occlusion treatment.
ATS 2 is ongoing, and consists of three randomised controlled trials (RCTs) of occlusion dosage. In moderate amblyopia (20/40–20/80), 2 hours per day is compared with 6 hours per day of occlusion. In severe amblyopia (20/100– 20/400) 6 hours per day is compared with full time occlusion. The incidence of amblyopia recurrence will be observed for 12 months after treatment cessation. This will lead to a RCT of maintenance occlusion treatment versus no treatment for the prevention of amblyopia recurrence. The methodology of ATS 2 is particularly open to the criticism that compliance monitoring is relatively weak.
The RCT phase of ATS 3 has recently commenced (Jonathan Holmes, personal communication). This is a study of older children found to have amblyopia of 20/40–20/400. One trial compares spectacles alone and spectacles plus occlusion plus atropine penalisation in 7–12 year olds. A second trial compares spectacles alone and spectacles plus occlusion in 13–17 year olds.
ATS 4 also commenced recently (Jonathan Holmes, personal communication). This is a RCT of weekend atropine penalisation versus daily atropine penalisation in moderate amblyopia.
The results of these studies, along with additional studies derived from their methods, will soon allow evidence based treatment of amblyopia for the first time—300 years after occlusion treatment was introduced. What will Tan et al find in 5 years’ time, if they repeat their study? Firstly, atropine penalisation may be the first line treatment for the majority of children with amblyopia. Secondly, and more importantly, amblyopia treatment advice given by ophthalmologists and orthoptists around the world will be consistent, and evidence based.
Note in Proof
More evidence that we need evidence