The Institutional Review Board of Indiana University approved this research protocol and the study adhered to the tenets of the Declaration of Helsinki and to the Health Insurance Portability and Accountability Act (HIPAA). 

Study population

We enrolled children between the ages of 3 and 8 years with untreated, moderate, unilateral amblyopia, diagnosed and cared for by pediatric ophthalmologists at Indiana University School of Medicine. Informed consent was obtained from the subjects’ parent or guardian (hereafter referred to as “parent”); assent was obtained from subjects 7 to 8 years of age.

Eligibility testing included the measurement of bestcorrected visual acuity (VA) in both eyes using the standard ATS single-surround HOTV letter protocol ⁽¹⁸⁾ and a routine comprehensive eye exam (cycloplegic refraction, comprehensive ocular examination and a full motility examination). 

Eligibility inclusion criteria: (I) ages range from 3 to 8 years old; (II) unilateral amblyopia: best corrected VA of the amblyopic eye ranging from 20/40 to 20/160. VA in the sound eye at least 20/40 or better. Interocular logMAR difference is at least two logMAR lines; (III) amblyopia associated with strabismus, anisometropia, or both; (IV) wearing of optimal spectacle correction (if needed) for a minimum of 12 weeks prior to enrollment. Details of the protocol for correction of refractive error followed guidelines from a previous PEDIG amblyopia treatment study of moderately amblyopic children ⁽¹⁹⁾; (V) previous patching treatment is allowed.

Ten children with amblyopia (20/40–20/125 in the amblyopic eye) were enrolled. A total of seven came back for 4-week follow-up, and six of them had 4-week VA measurements. The VA of one child was not measured because of atropine eye drop application in the morning of the visit. At the 12-week follow-up, six of the subject’s visited back, and unfortunately only three of them brought the MEMS bottle back. Therefore, while we present an example of the 12-week follow-up, we focus on reporting data from the 4-week follow-up. The baseline characteristics of the patients are shown in Table 1.

Table 1 Baseline characteristics of the patients who had follow-up visits

Example of objective compliance with atropine penalization

Figure 1 shows both the objective and subjective compliance from an individual (IW) followed for 12 weeks. The overall average objective compliance for this patient (IW) was 57% while subjective compliance was 86%. As Figure 1 clearly shows, subjective compliance is not enough to provide information about the detailed dosage over time. After the 8th week, the objective compliance of this patient was generally lower than 50%.

Figure 1 Both objective and subjective compliance over 12 weeks in a 6-year-old child with a daily regimen.

This pilot study demonstrates that objective compliance with atropine penalization can be monitored with MEMS; objective compliance shows a trend for decreasing compliance over time and with increasing regimen frequency. Using the MEMS approach may facilitate more effective communication between clinicians and patients. Monitoring compliance objectively may provide further understanding of the patients’ parents’ compliance behavior and will lead to more precise and accurate dosage.

Although this was a pilot study with a small sample size, we have shown that objective compliance with atropine penalization is not always as high as expected or as patients’ parents report. It varies with individuals and is associated with regimen. This indicates that prescribing higher regimen may not be as eff ective as expected because patients may not comply as well with a regimen that requires more frequent drop instillation. In addition, instructions for instilling eye drops in patients’ eyes is often vaguely described on the commercial eyedropper; it says “instill 1–2  drops”. Thus, we may not know the actual dosage applied to patients. 

As Figure 3C shows, although prescribed with a more frequent regimen, patient IW did not actually instill a higher dosage of eye drops compared to patient CT, who was prescribed with a less frequent regimen. Such findings suggest that we cannot merely count on prescribed regimen. Our pilot data showed that the actual dose in gram versus response relationship is a better fit. Although it is not robust to conclude with such a limited sample size, our approach demonstrated the potential to probe the actual dose in grams versus response relationship with more information from quantitative measurement. 

In addition to the small sample size, this study is limited by a few assumptions: (I) we assumed that one opening of the bottle corresponded to one instillation of eye drop; (II) we assumed that a decreasing eye drop bottle weight corresponds to the volume of atropine instilled into the eye. Even given these assumptions, the objective estimate of compliance and amount of eye drops still provides more quantitative information than previous studies. 

Atropine penalization was recommended as a firstline treatment for amblyopia, but it has not been widely applied in clinical settings ^(21,22). One reason could be that there is no study to clarify the dose-response relationship in atropine penalization treatment. The findings from a previous study indicate that a daily regimen (7 drops/ week) of atropine penalization is equally effective as a weekend regimen (2 drops/week) for both moderate and severe amblyopia ⁽²³⁾. However, the similar outcomes between the daily and weekend cohorts may have resulted from differences in patient compliance. Without objective compliance measurement, it is impossible to know whether the daily regimen group instilled the drops on a daily basis. To date, no clear evidence exists regarding the doseresponse relationship between atropine treatment and visual outcome. Thus, our pilot study points to the importance of deff ning a clear dose-response relationship and assessing objective compliance to atropine penalization.

An investigation of compliance and further doseresponse of atropine treatment is critical for the following three reasons: (I) as a pharmaceutical, atropine may have more serious side effects than the adhesive patch. Thus, it is critically important to find the minimal atropine dose needed to achieve effective vision improvement; (II) unlike patching, which can be immediately removed, atropine affects accommodation, pupil size, and near VA for more than 72 hours ⁽²⁴⁾. As has been previously reported, atropine treatment has a risk of inducing reverse amblyopia (degraded vision in the treated eye) if not monitored closely ^(25,26). Thus, the question of when to cease atropine treatment is pivotal; (III) atropine may take a longer time to be effective than patching. In a randomized clinical trial conducted by PEDIG on patching and atropine, after 5 weeks 56% of patients were successfully treated in the patching group and only 33% in the atropine group; by 6 months, these percentages were similar (79% and 74%, respectively) ^(3,7). This finding indicates that atropine treatment requires compliance for a longer period. At this point, we have very little information about the doseresponse relationship and more experiments with longer follow-up should be designed to explore these questions. 

Conclusions