Accuracy of a hand-held autorefractor in a pediatric population

Purpose. To evaluate a hand-held autorefractor in children for measurement ability and accuracy, as compared to cycloplegic retinoscopy. Methods. The Retinomax (Nikon Corp, Melville, NY) was used by two clinicians on two groups of subjects. The right eye of each subject was analyzed. Group 1 consisted of 47 subjects between 12 and 93 months of age, while Group 2 consisted of 11 subjects between 8 and 16 years. Immediately following cycloplegic retinoscopy (and subjective refinement, if possible), autorefraction was attempted. In all but the youngest subjects, 8 estimates of refractive error were made by the Retinomax and a composite measurement determined. Results. Group 1 had more astigmatism (mean 1.23 D) than Group 2 (0.18 D). Measurements could be obtained on seven of nine subjects less than 36 months of age, and on all older subjects The vector dioptric difference, taking into account sphere, cylinder and axis differences simultaneously, was 0.84 +/- 0.53 (X̄ +/- sd) diopters in Group 1, and ranged from 0.1 to 2.54 diopters In Group 2, vector differences averaged 0.44 +/- 0.29 D (X̄ +/- sd), and ranged from 0.0 to 0.89 diopters The subject with the greatest difference (2.54 D) also had the largest refractive error (-8.00 +6.75 ×080). Conclusions. The Retinomax was readily used in all subjects older than three years, and with nearly all subjects between one and three. The agreement between the Retinomax and retinoscopy is comparable to that between two independent retinoscopists (Safir, 1970).