Concordance with occlusion therapy for childhood amblyopia

Abstract

 

Purpose

 

The Monitored and Randomized Occlusion Treatment of Amblyopia Studies (MOTAS & ROTAS) used Occlusion Dose Monitors to measure objectively occlusion. Here we examine factors influencing concordance with prescribed occlusion dose.

 

Methods

 

Studies had a 3-phase protocol: initial assessment, refractive adaptation and occlusion. Occlusion therapy proceeded refractive adaptation for those with unresolved amblyopia. Participants were instructed to dose for 6 hrs/day (MOTAS) or randomized to 6 or 12 hrs/day (ROTAS). Occlusion was monitored at ≈1 min intervals during treatment, and dose data downloaded for analysis during 2-weekly follow-up clinic visits.

 

Results

 

152 patients (mean;sd age 65;18 months) received occlusion. Amblyopia was associated with anisometropia in 50, strabismus in 44, and both (mixed) in 58. Median follow-up was 99 days (IQR 72 days). Mean concordance was 44%, mean proportion of days with zero occlusion was 42%. Concordance at weekends was lower than on weekdays (p=0.039) as was the likelihood of dosing at all (p=0.028). Concordance was lower, and dosing on any given day less likely, the later into follow-up dosing occurred and with less frequent clinic visits (all p<0.001). Those prescribed 6 hours of occlusion had lower concordance (p=0.033) than those prescribed 12, but both groups were equally likely to dose at all on a given day. Age, gender, amblyopia type and severity were not associated with concordance. Among non-zero daily occlusion doses, mixture modelling indicated three underlying subpopulations: doses of usually under 10% concordance; doses of usually 30%-80% concordance; and doses of usually near-100% concordance.

 

Conclusions

 

Concordance is strongly influenced by the frequency of days with zero dose, which are more likely at weekends and increasingly frequent over time in follow-up. Minimizing such days should be a priority if attempting to boost concordance (e.g. if participants had received their mean non-zero daily dose on half of the days where they did not dose at all, overall concordance would have been 57% rather than 44%).

 

 

Mean daily concordance with prescribed dose across all individuals still in treatment. Days with fewer than 10 individuals not shown.

 

 

Bars: non-zero daily dose concordance distribution across all individuals. Lines: components of 3-part normal mixture model (fit to log-transform of data). Legend percentages: estimated proportion of daily doses in each component distribution.

 

Keywords: 417 amblyopia • 466 clinical (human) or epidemiologic studies: treatment/prevention assessment/controlled clinical trials • 473 computational modeling