Patching Regimens: Author Reply

Clinical trials have served as the foundation of much of the knowledge we have gained in the last 50 years with regard to the efficacy of treatments of various medical conditions. In very few conditions is the natural history of the disease so well defined, and the outcome of intervention so evident, that the benefits and adverse effects of a treatment can be determined without a prospective, controlled trial. Sometimes trial results confirm clinical impressions, and sometimes they do not. Drs von Noorden and Campos indicate that in their experience full-time patching produces better results than part-time patching in treating severe amblyopia. However, they provide no data to support this conclusion, and we are not aware of a prospective study that supports this claim. They indicate that, for most clinicians, full-time patching has been the standard of care for severe amblyopia. Nevertheless, whether to prescribe full-time or part-time patching for severe amblyopia is a controversy recently reported in the literature1Tan J.H Thompson J.R Gottlob I Differences in management of amblyopia between European countries.Br J Ophthalmol. 2003; 87: 291-296Crossref PubMed Scopus (25) Google Scholar and confirmed by a survey of our investigator group in 1998. Drs von Noorden and Campos state that “children who wear their patch only part of the prescribed time do not respond as well to treatment,” citing studies that report the response to patching, as a function of compliance, and measured by an occlusion dose monitor.2Loudon S.E Polling J.R Simonsz H.J A preliminary report about the relation between visual acuity increase and compliance in patching therapy for amblyopia.Strabismus. 2002; 10: 79-82Crossref PubMed Scopus (49) Google Scholar, 3Loudon S.E Polling J.R Simonsz H.J Electronically measured compliance with occlusion therapy for amblyopia is related to visual acuity increase.Graefes Arch Clin Exp Ophthalmol. 2003; 241: 176-180Crossref PubMed Scopus (71) Google Scholar In these studies, 14 children were prescribed a wide range of part-time occlusion regimens (from 15 minutes a day to 3 complete days per week) for a wide range of amblyopia severity (0.78 [20/25]–0.03 [20/667]). We question whether these data are useful in addressing the issue of part-time versus full time patching in severe amblyopia. In addition, any analysis of clinical trial data limited to compliant patients has considerable potential for bias and production of erroneous conclusions, particularly in a condition such as amblyopia, in which the response to treatment may influence subsequent compliance. There are examples in the literature of erroneous conclusions drawn from selected compliant populations.4Chene G Morlat P Leport C et al.Intention-to-treat vs. on-treatment analyses of clinical trial data experience from a study of pyrimethamine in the primary prophylaxis of toxoplasmosis in HIV-infected patients.Control Clin Trials. 1998; 19: 233-248Abstract Full Text Full Text PDF PubMed Scopus (39) Google Scholar, 5The Coronary Drug Project Research GroupInfluence of adherence to treatment and response of cholesterol on mortality in the Coronary Drug Project.N Engl J Med. 1980; 303: 1038-1041Crossref PubMed Scopus (663) Google Scholar Drs von Noorden and Campos contend that a treatment group difference might have become apparent with longer follow up. In designing the trial, the 4-month follow-up period represented the maximum length of time we believed that the fixed treatment regimens (6 hours/day or full-time patching) could be prescribed before a change in the treatment might be necessary for patients who had not responded well to the randomized treatment assignment. It is conceivable that treating for longer might show greater benefit with full-time than part-time patching, although there is no suggestion of even a trend in our data favoring a delayed benefit for full-time patching comparing our results at 5 weeks with our results at 4 months. As noted by Drs von Noorden and Campos, we state that our study compared prescribed patching regimens and not the actual number of hours of occlusion that were achieved. We commented in the “Discussion” that it is possible that the results reflect, at least in part, suboptimal compliance in the full-time patching group. However, we designed this trial to assess the effectiveness of the prescribed patching regimens rather than their efficacy. Efficacy refers to whether a treatment has benefit under highly controlled conditions, whereas effectiveness refers to whether a treatment has benefit as used in the real world. A greater number of hours of actual patching could have greater efficacy than fewer hours, but if this intensity of patching can only be achieved in a minority of cases, it would not be a very effective treatment. Our compliance data suggest that most patients are not able to achieve full-time patching, so although it is of scientific interest as to whether full-time occlusion can produce a better outcome than part-time occlusion, this may have limited clinical utility. We have not yet studied the potential benefit of full-time patching in patients who have had an incomplete response to part-time patching, but our data suggest it is entirely reasonable to prescribe an initial course of part-time occlusion for severe amblyopia. In response to Dr Greenberg's comments, the inclusion of multiple community-based sites, rather than limiting a trial to tertiary referral centers, has the advantage of enhancing generalizability of the results. The type of patient enrolled in this study, with severe anisometropic or strabismic amblyopia and no patching during the preceding 6 months, was seen much more commonly at our community-based sites than at our institution-based sites. Including multiple community-based centers in the trial design is not unorthodox, but is a priority for future clinical trials funded by the National Institutes of Health, according to its director.6Re-engineering the clinical research enterprise. Available at: http://www.nihroadmap.nih.gov/clinicalresearch/index.asp. Accessed January 13, 2004Google Scholar As a rule, it is highly inaccurate to extrapolate from an individual's recollection of his or her own practice to the number of patients that can be recruited for a randomized clinical trial with strict inclusion criteria such as severe anisometropic or strabismic amblyopia and no patching during the preceding 6 months. Clinical trialists often use 50% or even 10% of the number cited by individual investigators before a study when considering feasibility of recruitment. As noted above, a substantial proportion of pediatric eye care providers prescribed part-time patching for severe amblyopia before the initiation of our study. Many eye care providers request that the child perform some activities at near requiring visual attention for some of the time the patch is worn. Because, as noted by Dr Greenberg, it is not known whether performing near activities influences the response to occlusion therapy, this was included in the protocol in both groups to standardize that aspect of the treatment regimens. We will be conducting a future trial to determine whether near activities are beneficial during patching. Dr Greenberg provides a single case report to refute our findings. Of interest, in the occlusion dose monitor studies cited earlier,2Loudon S.E Polling J.R Simonsz H.J A preliminary report about the relation between visual acuity increase and compliance in patching therapy for amblyopia.Strabismus. 2002; 10: 79-82Crossref PubMed Scopus (49) Google Scholar, 3Loudon S.E Polling J.R Simonsz H.J Electronically measured compliance with occlusion therapy for amblyopia is related to visual acuity increase.Graefes Arch Clin Exp Ophthalmol. 2003; 241: 176-180Crossref PubMed Scopus (71) Google Scholar there was one patient with severe strabismic amblyopia and an initial visual acuity (VA) of 20/200 who improved to 20/25 with only 1 hour of patching per day. We do not believe that a conclusion should be drawn from the latter case, that little patching improves severe amblyopia, any more than a conclusion should be drawn from Dr Greenberg's case that full-time patching is needed to treat successfully every case of severe amblyopia. Disparate results in case reports and case series point to the critical importance of prospective, randomized trials for providing meaningful, clinically relevant information for establishing treatment guidelines. Randomization reduces the potential selection bias to which Dr Greenberg refers. Complete masking of all outcome measurements is rarely obtainable in clinical trials, and >90% masking would be considered excellent. In our study, the VA outcome was measured using a highly automated, computerized, VA testing protocol, where there was little opportunity for even a biased nonmasked tester to influence the results.7Holmes J.M Beck R.W Repka M.X et al.The Amblyopia Treatment Study visual acuity testing protocol.Arch Ophthalmol. 2001; 119: 1345-1353Crossref PubMed Scopus (257) Google Scholar, 8Moke P.S Turpin A.H Beck R.W et al.Computerized method of visual acuity testing; adaptation of the Amblyopia Treatment Study visual acuity testing protocol.Am J Ophthalmol. 2001; 132: 903-909Abstract Full Text Full Text PDF PubMed Scopus (202) Google Scholar We believe that our trial of 175 patients provides the best evidence available to date on the response of severe amblyopia to different prescribed patching regimens. We contend that our study was conducted with a high degree of rigor and that the validity of the results are comparable to that of any other trial funded by the National Eye Institute, including the 2 trials Dr Greenberg mentions. Perhaps Dr Greenberg will be reassured in knowing that an independent data and safety monitoring committee (Drs Barlow, Buckley, Davis, Dobson, Keltner, Osman, Palmer, and Phelps—see article for listing of full names and affiliations) appointed by the National Eye Institute approved the protocol before its initiation, monitored the data during the conduct of the trial, reviewed the results at the end of the trial, and approved the manuscript before its submission to Ophthalmology. The Pediatric Eye Disease Investigator Group, which conducted this trial, remains open to new investigators. We would welcome the involvement of Dr Greenberg and others with busy pediatric eye care practices. For information on becoming an investigator in the Pediatric Eye Disease Investigator Group, please contact the Pediatric Eye Disease Investigator Group Coordinating Center at [email protected].