Abstract
Purpose: :
To determine if the astigmatism accuracy and measurement range of the SureSight Autorefractor (Welch Allyn, Inc) is improved by overrefraction through a correcting lens in a population of Native American children having a high prevalence of astigmatism.
Methods: :
Subjects were 24 2nd and 3rd Grade students who attended school on the Tohono O’odham reservation. The children had right eye non–cycloplegic autorefraction performed with the Retinomax Plus (Right Mfg, Inc) to provide a "gold standard" astigmatism measuremetn, followed by 3 non–cycloplegic readings with the SureSight. The median of the 3 SureSight measurements was taken as an initial estimate of cylinder and axis. A lens containing the initial estimate of cylinder from the SureSight was used for overrefraction: the lens was then handheld before the subject’s eye and 3 additional SureSight readings were taken through the correcting lens. When no reading was obtained on the first set of readings, a +4 D cylinder lens at axis 90 degrees was used as the correcting lens. The median cylinder and axis from the 3 overrefraction measurements were added (in vector form) to the results of the initial estimate to determine a final estimate of cylinder and axis. Vector differences (VDD) between the Retinomax estimate of cylinder and the initial and final SureSight estimates were calculated.
Results: :
For 4 of 24 subjects (all of whom had astigmatism > 3.00 D), no reading could be obtained with the SureSight autorefractor. By refracting through the handheld lens, an estimate could be obtained on all subjects. For the 20 subjects on whom we could obtain both an initial and a final measurement of cylinder, the final estimate (calculated as the vector sum of the handheld lens power and the SureSight measuremetn through the lens) was closer to the Retinomax measurement than was the initial estimate on 15/20. For all 24 subjects, the mean VDD between the overrefraction estimate with the SureSight and the Retinomax estimate of astigmatism was 0.42D (SD 0.28D). For the 4 subjects on who we could not obtain initial estimates of cylinder, mean difference between the final estimate and the Retinomax estimate was 0.65 D (VDD) (SD=0.01D).
Conclusions: :
Testability and accuracy of SureSight autorefractor measurements of cylinder are improved by overrefraction based on an initial estimate of cylinder obtained with the SureSight. The overrefraction protocol allows estimation of astigmatism children with levels of astigmatism beyond the range of the SureSight.
Keywords: astigmatism • refractive error development • clinical (human) or epidemiologic studies: prevalence/incidence