Patients and Methods
A 10-question, multiple-choice, close-ended questionnaire (Table A, available in the online version of this article) was developed and distributed to all pediatric ophthalmologists and fellows attending the Annual Joseph H. Calhoun Pediatric Ophthalmology Forum at Wills Eye Hospital on November 4, 2016.
Anonymous Survey: Practice Patterns in the Management of Amblyopia
The questionnaire consisted of basic demographic information (questions 1, 2, and 3) and elicited preferences in management of amblyopia or amblyopia risk factors under different clinical scenarios.
Respondents were allowed to skip questions they did not want to reply to and, for some questions, were permitted to choose an “other” option whenever the provided options did not match their policies. Surveys were excluded if filled out by anyone other than a fellow or attending physician.
The surveys were completely anonymous and did not contain any identifying information. The Internal Review Board Committee of the Wills Eye Hospital, Philadelphia, Pennsylvania, considered this study to be exempt from internal review board review and approval.
Survey responses were summarized using measures of frequency and percentages. To assess for a relationship between respondent demographics and survey responses, a two-tailed Fisher's exact test of independence was performed with an alpha value of 0.05 for all statistical analyses.
The questionnaire was given to an estimated 113 pediatric ophthalmologists and fellows at the meeting. A total of 74 (65.5%) completed the survey, although only 73 responded for question 9, 72 for question 6, and 68 for questions 2 and 3. Of the respondents, 92% (68 of 74) were attending physicians and 8% (6 of 74) were fellows. Of the 68 attending physicians, 31% (21 of 68) had been in practice for less than 10 years and 70% (47 of 68) had been in practice for more than 10 years; 37% (25 of 68) were academic or university affiliated and 63% (43 of 68) worked in a private practice setting.
There was no statistically significant relationship (P < .05) between any of the variables, indicating that no group (eg, fellow vs attending, academic vs private practice, years in practice) was more or less likely to respond to the question in any way (Table B, available in the online version of this article).
Survey Response Frequencies and Percentages
Seventy-six percent (56 of 74) of respondents replied that they would prescribe refractive correction to a 3-year-old child with untreated anisometropic amblyopia prior to initiating occlusion therapy. In contrast, 23% (17 of 74) would prescribe refractive correction and part-time occlusion therapy together. Only 1% (1 of 74) would prescribe refractive correction and full-time occlusion.
The preference to prescribe refractive correction prior to patching was similarly reported by attending physicians and fellows and pediatric ophthalmologists with more than 10 years of experience and those with fewer years of experience. However, there was variation between those practicing in different settings, with 28% (12 of 43) of clinicians in private practice choosing combined therapy initially, compared to only 12% (3 of 25) of those practicing in an academic setting (Table 1).
Preferred Management of a 3-Year-Old Patient With Untreated Anisometropic Amblyopia Divided by Training Level, Length of Practice, and Practice Setting
Timing of Strabismus Surgery for Exotropia
In a child with exotropia and coexisting amblyopia, 57% (42 of 74) replied that they would prescribe refractive and occlusion therapy until significant visual improvement was achieved prior to performing strabismus surgery. In contrast, 29.7% (22 of 74) would prescribe refractive and occlusion therapy, performing strabismus surgery even with only mild improvement in visual acuity. An additional 13.5% (10 of 74) chose “other,” with one respondent writing that it depended on the appearance of the exotropia and another replying that he or she would perform surgery after the visual acuity plateaued.
The degree of variation among clinicians on timing of strabismus surgery was fairly consistent among those with different lengths of practice and practice settings (Table 2). However, the difference was more pronounced between fellows and attending physicians, with 83% (5 of 6) of fellows waiting for significant visual improvement before performing strabismus surgery compared to 56% (38 of 68) of attending physicians.
Timing of Strabismus Surgery for Exotropia by Pediatric Ophthalmologists Divided by Training Level, Length of Practice, and Practice Setting
Method of Patching
Seventy percent (52 of 74) of respondents preferred using an opaque adhesive applied directly to the skin, whereas 23% (17 of 74) preferred using a patch placed on an eyeglass frame. No respondents chose opaque contact lenses. An additional 7% (5 of 74) replied “other,” all of whom indicated no preference between an opaque adhesive applied directly to the skin or a patch placed on an eyeglass frame. Method of patching was not different based on practice setting or between fellows and attending physicians. However, there was a difference based on practice duration; 81% (17 of 21) of clinicians in practice less than 10 years preferred an opaque adhesive applied to the skin compared to 66% (31 of 47) of those in practice for more than 10 years.
For a 4-year-old patient with moderate anisometropic amblyopia, 86% (63 of 73) replied that they would prescribe part-time patching, whereas only 10% (7 of 73) responded that they would use atropine. An additional 4% (3 of 73) reported that they would prescribe full-time patching. Pediatric ophthalmologists of all levels of training and settings similarly reported a preference for part-time patching. One respondent did not answer this question.
Cessation of Therapy
Fifty-seven percent (41 of 72) reported that they would stop occlusion therapy at 10 years of age or older, 24% (17 of 72) at 9 years of age, and 19% (14 of 72) at 7 to 8 years of age. One respondent wrote that it depended on when the patient started occlusion therapy, whereas another replied that it depended on the patient's response to initial therapy. Two respondents (2%) did not answer the question.
This was similarly reported by pediatric ophthalmologists with 10 or more years of experience and those practicing in academic or private practice settings. However, among fellows, 50% (3 of 6) would stop occlusion therapy at 10 years of age or older, 50% at 9 years of age, and no respondents (0 of 6) at 7 to 8 years of age.
Treatment of Hyperopic Anisometropia During the First 1.5 Years of Life
For a 15-month-old patient with a 1.50-diopter hyperopic anisometropia, 51% (38 of 74) of respondents reported that they would prescribe no refractive correction with follow-up in 6 months, whereas 15% (11 of 74) would prescribe no refractive correction with follow-up in 1 year. Twenty-four percent (18 of 74) would prescribe spectacles with partial cycloplegic refractive correction, 7% (5 of 74) would prescribe spectacles with no patch, and 3% (2 of 74) would prescribe spectacles with full cycloplegic refractive correction.
The preference for prescribing refractive correction was not different based on years in practice or practice setting. However, a difference did exist between fellows and attending physicians in regard to length of follow-up when not prescribing refraction; 33% (2 of 6) of fellows reported that they would follow up in 6 months and 33% (2 of 6) in 1 year compared to 13% (9 of 68) of physicians who would follow up in 6 months and 52% (35 of 68) in 1 year.
Estimated Patching Compliance
The estimated patching compliance rate was 50% to 75% for 64% (47 of 74), 75% to 100% for 19% (14 of 74), and 25% to 50% for 18% (13 of 74). No respondents reported a compliance rate of less than 25%. This did not differ based on length of practice or practice setting. However, no fellows estimated a compliance rate of 75% to 100%, whereas 50% (3 of 6) reported a compliance rate of 25% to 50% compared to 21% (14 of 68) and 15% (10 of 68) of attending physicians, respectively (P = .08). A trend toward higher estimated compliance was also noted in private practice settings compared to academic settings (Table 3).
Estimated Patient Compliance for Patching by Pediatric Ophthalmologists Divided by Training Level, Length of Practice, and Practice Setting
Treatment of amblyopia has existed for centuries; however, there are still few guidelines that standardize treatment. Previous studies have found significant variation in amblyopia management among providers at different centers and countries, but did not assess for the physician-related variables explored in our study.12–14 A retrospective multicenter study by Mazow et al.12 found statistically significant physician variation in method of treatment termination, patient compliance, surgery performed, and duration of therapy based on center but not age or sex. Loud-on et al.14 found that Dutch and German orthoptists varied widely in their prescription hours of occlusion therapy on a case per case basis, whereas Tan et al.13 found that orthoptists in German-speaking countries prescribed more hours per week than those in the United Kingdom.
Our study demonstrated both consistencies and variations in question responses overall, which is likely multifactorial but could be related to the presence or lack of practice guidelines and randomized clinical trials, respectively. We found no statistically significant relationship between any group demographic and their response to treatment questions. This suggests that management decisions are not influenced by level of training, years in practice, or practice location or could be influenced by other demographic factors. It is also possible that the sample size was too small to detect a relationship. Clinical decision-making may be purely influenced by one's own interpretation of the current guidelines and results of randomized controlled trials as they become available.
The majority of respondents (75.5%) showed a preference for initiating refractive correction alone prior to occlusion therapy in patients. This is in agreement with the most recent Preferred Practice Patterns publication by the Academy of Ophthalmology, which states that treatment of refractive error alone can improve in children who have untreated anisometropic or strabismic amblyopia. This is also supported by multiple randomized controlled trials that have demonstrated improvement or resolution of amblyopia from refractive correction alone.5,6,17–19 No studies have explored the influence of demographic factors on this aspect of management, whereas our study found no relationship.
Similarly, the preference for part-time patching among respondents (86.3%) is in line with a series of well-known randomized controlled trials that found that fewer prescribed hours of occlusion therapy or penalization with atropine is just as effective over the same treatment period in patients with both moderate and severe amblyopia.20–22 This is in contrast to a previous survey that found that the majority of pediatric ophthalmologists did not alter their clinical practice after the publication of the Pediatric Eye Disease Investigator Group patching trials.17 The response to this question did not vary among demographic groups, whereas previous studies found significant variations in patching regimens among different European countries but did not assess for other demographic variables.13,14
Occlusion therapy is largely dependent on compliance, which can be difficult to achieve in children and varies from practice to practice.12,23 More than half of respondents estimated a patching compliance of 50% to 75% in their practice. There was no significant relationship between respondent characteristics and compliance rates, although a larger percentage of fellows (50%) reported poorer compliance than attending physicians, a difference that approached statistical significance (P = .08). Mazow et al.12 reported a significant variation in estimated compliance rates among pediatric ophthalmologists at nine different centers around the United States (P < .0001), but did not report on respondent demographics.
One of the more controversial aspects of amblyopia treatment is the management of patients with strabismus and coexisting amblyopia. Current guidelines recommend that amblyopia treatment be started prior to surgery but do not specify timing of surgery.18,24,25 More than half of respondents (56.8%) would begin amblyopia therapy until significant improvement in visual acuity and then proceed with strabismus surgery, whereas 29.7% would prescribe amblyopia therapy, proceeding with strabismus surgery even with only mild improvement in visual acuity. No one group was more likely to choose one answer over another. Variation among different centers has been reported on surgery alone versus surgery with continued patching, but no previous study has commented on physician-related variables on optimal timing of surgery.12
Most responses agreed with the classic teaching that amblyopia be corrected to maximum result prior to realignment surgery; however, this concept has recently been refuted.12,26 Lam et al.26 found that corrective surgery in children with esotropia before full resolution of amblyopia is not only safe and efficient if the amblyopia therapy is continued after surgery, but also suggested that surgery alone can reverse amblyopia in some cases. Similarly, Dadeya and Kamlesh9 concluded that it is not mandatory to treat amblyopia prior to surgical correction unless it is a case of infantile esotropia with moderate amblyopia.
Another controversial aspect of treatment is the influence of age on cessation of treatment. The current Preferred Practice Patterns publication by the Academy of Ophthalmology recommended that treatment should be offered to children regardless of age; however, there are no validated studies to establish when therapy can be discontinued. Common practice is to require three consecutive visits separated by 6 to 8 weeks without improvement in visual acuity in the setting of therapy compliance and attempts with both occlusion and penalization therapy.27 The majority of respondents (56.9%) would stop occlusion therapy at 10 years or older or until better visual acuity was achieved, whereas the remainder would cease treatment at a younger age. This likely reflects recent literature indicating that therapy may be effective past the previously identified “sensitive period” of 7 years of age, especially in those not previously treated.28,29
This question received the most comments and omitted answers, with requests for additional information such as if and when treatment was initiated. This suggests that the age of the patient is becoming more controversial and that course of treatment and duration of time without improvement are better indicators for making the decision to end treatment. There are no previous studies that report on the influence of demographic variables on age of cessation of treatment. However, one study found significant practice variations on the manner in which treatment was ended (abruptly, tapered, or maintained) among different centers. This indicates that the age of cessation of treatment is less clear and likely both case dependent and related to physician preferences.
Anisometropia, a well-known risk factor for amblyopia, is another controversial element of amblyopia treatment and prevention. The decision to prescribe correction is dependent on the type of refractive error, the severity, and the child's age. Our survey presented respondents with a 15-month-old patient with a hyperopic anisometropia of +1.50 diopters. The majority of respondents (66.3%) would not prescribe spectacles with follow-up at either 6 months or 1 year, whereas 33.8% would prescribe refractive correction. No one group was more or less likely to prescribe correction over the other and there are no studies evaluating the influence of physician demographic factors.
Current Preferred Practice Patterns by the Academy of Ophthalmology recommend correction for ages 1 to 2 years with a hyperopic anisometropia of +2.00 diopters or greater; however, the threshold would be lowered in the presence of amblyopia.18 These guidelines create a sizable gray area because detection of amblyopia in children is more difficult due to illiteracy and inability to read eye charts, which may explain why other respondents chose to prescribe refractive correction.
Limitations of our study include a small sample size and failure to identify other factors that may play a role in practice patterns (eg, sex or geographic location). Distribution of a similar survey at conferences in different locations would not only permit a larger sample size, but would also allow evaluation of geographical variables on practice patterns.
Our results suggest that practice patterns appear to be more uniform among pediatric ophthalmologists when there are clearly published guidelines and well-known randomized controlled trials that address clinical management. In contrast, there may be more variation in clinical practice when there are no clear published guidelines or randomized controlled trials are lacking. Our study showed no statistically significant relationship between responses and respondent characteristics, indicating that number of years in training and practice location did not influence treatment decisions in the survey questions.
This study emphasizes the need to not only identify practice pattern variations among pediatric ophthalmologists in the treatment of amblyopia, but also the need for well-designed prospective randomized controlled trials to establish treatment guidelines and determine whether practice approaches standards of care.