The Pediatric Eye Disease Investigator Group Report May Be Too Optimistic About Efficacy of Treatment: In Reply

Abstract

In Reply.—Dr Lempert's recent letter describes similar issues to those he has raised in previous letters to the editor. We have responded to these issues in the past and will reiterate our responses. The Amblyopia Treatment Studies (ATS) are a series of randomized, clinical trials designed to answer questions about the optimal management of amblyopia.1–9 The studies are conducted by the Pediatric Eye Disease Investigator Group (PEDIG), a large group of pediatric ophthalmologists and optometrists based in both university and community settings.10,11 Dr Lempert contends that a significant weakness of the ATS is the inclusion of only a small number of subjects from each center and suggests that this reflects a selection bias. In fact, by involving a larger number of participating centers, we are able to achieve a sample size sufficient to make statistically valid observations more rapidly than could be accomplished if only a limited number of centers participated. In addition, including a large number of diverse centers increases generalizability of the results.We agree that a clinical trial can never completely reflect daily practice. The informed-consent process itself probably educates parents of amblyopic children more than the usual instructions provided during routine care, and this may affect treatment compliance. Nevertheless, we deliberately designed the ATS to be conducted in a “real-world” environment as much as possible, with the exception of randomization of treatment assignment and a standardized protocol for assessing visual acuity.Dr Lempert contends that an additional weakness of the ATS was the lack of an untreated control group. When we designed our early studies, PEDIG investigators concluded that they could not maintain clinical equipoise (a balance between potential benefit and potential harm between treatment groups) in trials that required maintaining the child in an assigned untreated group for 4 to 6 months. However, in our analysis of these early studies, we did not find a significantly different treatment effect in older children, compared with younger children, within our 3- to 7-year age range. These new data have allowed us to design a short-duration study incorporating an untreated control group. We are currently conducting that randomized trial, comparing visual acuity improvement over 5 weeks between an untreated group and a group assigned to 2 hours of daily patching.Dr Lempert suggests that the observed improvement in visual acuity could occur because of maturation or a learning effect. In our completed ATS, the average improvement of amblyopic eye acuity ranged from 2.4 to 4.8 lines6,9 over 4 to 6 months. This improvement far exceeded that expected from a learning effect. It is important to note that the visual acuity improvement in amblyopic eyes also far exceeded the improvement seen in the fellow sound eyes over the same time period.Dr Lempert raises the possibility that organic causes may explain many cases of amblyopia observed in clinical studies. Although we cannot exclude rare cases of organic visual loss within our studies, we note that many of the amblyopic eyes in our studies improved to within the normal range. Thus, the presence of an abnormality of the optic nerve does not negate the importance of successful amblyopia treatment. Additional longer-term studies are needed to define whether these eyes become truly normal and, if so, whether they remain normal. We are currently conducting examinations on patients who were enrolled in our study of atropine versus patching and have reached the age of 10 years. We also plan to reexamine these patients at 15 years old.Dr Lempert suggests that we include biometry and disk imaging in future ATS. Using such technology would limit the number of centers that could participate in the studies and also exclude from participation a large number of the younger children who can provide recognition acuity and therefore participate in a randomized treatment trial but who would be unable to cooperate for biometric studies or fundus photography. Such morphologic studies are warranted but do not fit into our study design of large simple trials.Finally, Dr Lempert suggests that we have accepted “occlusion as the unquestioned mainstay of treatment for unilateral poor vision.” This is not true. However, in the past, we noted that patching is the mainstay of amblyopia therapy in North America. PEDIG studies have already investigated use of a pharmacologic agent, topical atropine, in the treatment of amblyopia.3 Thus far, we have concluded that both atropine and patching are appropriate for initial treatment of moderate amblyopia. Future randomized, clinical trials are needed to investigate additional modalities of amblyopia treatment in current use, including optical penalization, blurring filters, and systemic drugs.