Clinical comparison of the Welch Allyn SureSight handheld autorefractor versus cycloplegic autorefraction and retinoscopic refraction

J AAPOS. 2004 Apr;8(2):123-7. doi: 10.1016/j.jaapos.2003.10.006.


Purpose: To investigate the accuracy of the Welch Allyn SureSight (WASS; Welch Allyn, Skaneateles Falls, New York) hand-held autorefractor in noncyclopleged children and to compare the results with those obtained in cyclopleged children using the Nidek AR-820 (NAR; Nidek, Aichi, Japan) and retinoscopic refraction (RR) of an experienced pediatric ophthalmologist.

Methods: Ninety-one children (mean age 97 months +/- 19, range 37 to 107) whose vision screening, performed by a school nurse and/or by their primary care provider, was unsuccessful were prospectively evaluated in the private practice of a university pediatric ophthalmologist. After completion of a history, each child was refracted using the WASS. A complete ophthalmic examination, including cycloplegic refraction using the NAR and retinoscopic refraction refined to obtain a visual acuity of at least 20/30 diopters, was then performed. The physician was masked to the previous WASS and NAR results. Analysis variables included race, age (months), refraction (sphere, cylinder, axis), and initial and final corrected visual acuity. Only WASS values for which reliability was > or = 8 were used. A coin toss determined which eye would be used for analysis.

Results: Using analysis of variance and Scheffé multiple comparison testing, sphere results obtained by the WASS indicated a statistical difference compared with those obtained using the NAR (P =.0027) and retinoscopy (P =.0088). Similarly, spherical equivalence results obtained by the WASS also indicated a statistical difference compared with that of the NAR (P =.0027) and retinoscopy (P =.0056). Myopic sphere and spherical equivalent values were recorded more frequently with the WASS than with the NAR or refined retinoscopy. Cylindrical values comparing the WASS with the NAR, the WASS with RR, and the NAR with RR were statistically similar to each other (P =.9993,.748, and.7261 respectively). Axial results for patients with +0.50 or more of astigmatism were also statistically similar by all 3 methods.

Conclusions: Our data indicate that when the WASS is used on noncyclopleged children as suggested for primary care providers, myopia may be overdiagnosed.

Publication types

  • Clinical Trial
  • Controlled Clinical Trial
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Analysis of Variance
  • Child
  • Child, Preschool
  • Diagnostic Errors
  • Humans
  • Mydriatics*
  • Myopia / diagnosis
  • Prospective Studies
  • Refractive Errors / diagnosis*
  • Retinoscopy*
  • Vision Tests / instrumentation*


  • Mydriatics