Abstract
Purpose::
To determine the relation between age, astigmatism, and accuracy of noncycloplegic measurements of sphere obtained with the SureSightTM Autorefractor, as compared to cycloplegic measurements obtained with the Nikon Retinomax Autorefractor, in children between 2.6 and 7.6 years of age, from a population with a high prevalence of astigmatism.
Methods::
Subjects were 124 children enrolled in Head Start, kindergarten, or first grade during the 2005/06 school year, in schools located on the Tohono O’odham Reservation. Three right-eye, non-cycloplegic SureSightTM (SSNcR, Welch Allyn, child mode) measurements were attempted on each child. Cycloplegic (cyclopentolate 1%, 2 drops) autorefraction (RmaxCR) measurements were obtained with the Retinomax K+ (Nikon). SSNcR measurements were included only if the highest confidence measurement from the child had a confidence ≥ 6, and RmaxCR measurements were included only if they had a confidence ≥ 8. The difference between the highest confidence SSNcR (child mode) and the highest confidence RmaxCR measurements of sphere was determined for each subject. Mean differences for younger vs. older children (< 5 vs ≥ 5) were compared, as were mean differences for non-astigmats (< 1.00 D) vs astigmats (≥ 1.00 D).
Results::
Measurements were obtained for 112 children. The SSNcR measurements of sphere were approximately 0.75 D more plus (mean difference 0.73 D, SD 1.21, p < 0.001) than RmaxCR measurements of sphere. Difference in mean measurements of sphere did not significantly differ for younger (mean difference 0.76 D, SD 1.22) vs older children (mean difference 0.55, SD 1.16), nor did they differ for non-astigmats (mean difference 0.59 D, SD 1.01) vs astigmats (mean difference 0.90, SD 1.40). However, variability in measurement agreement between SSNcR and RmaxCR measurements of sphere across subjects was greater in astigmats than in non-astigmats.
Conclusions::
The SS (in child mode) tends to overestimate the amount of hyperopia in non-cyclopleged children in the 2.6 to 7.6-year old age range. The overestimation did not significantly differ by age or by presence of astigmatism. Greater variability in measurement agreement in astigmatic children suggests astigmatic children have more variation in their accommodation than do non-astigmatic children during SSNcR measurement.
Keywords: astigmatism • refraction