Noncycloplegic photorefractive screening in pre-school children with the "PowerRefractor" in a pediatric practice

Optom Vis Sci. 2007 Jul;84(7):630-9. doi: 10.1097/OPX.0b013e3180dc99ea.


Purpose: To provide a framework for typical refractive development, as measured without cycloplegia with a commercial infrared photorefractor. To evaluate the usefulness of the screening for refractive errors, we retrospectively analyzed the data of a large number of unselected children of different ages in a pediatric practice in Tuebingen, Germany.

Methods: During the standard regular preventive examinations that are performed in 80% to 90% of the young children in Germany by a pediatrician (the German "U1 to U9" system), 736 children were also measured with the first generation PowerRefractor (made by MCS, Reutlingen, Germany, but no longer available in this version). Of those, 172 were also measured with +3 D spectacles to find out whether this helps detect hyperopia. Children with more than +2 D of hyperopia or astigmatism, more than 1.5 D of anisometropia, or more than 1 D of myopia in the second year of life were referred to an eye care specialist. The actions taken by the eye care specialist were used to evaluate the merits of the screening.

Results: The average noncycloplegic spherical refractive errors in the right eyes declined linearly from +0.93 to +0.62 D over the first 6 years (p < 0.001)-between 1.5 and 0.5 D less hyperopic than in published studies with cycloplegic retinoscopy. As expected, +3 D spectacle lenses moved the refractions into the myopic direction, but this shift was not smaller in hyperopic children. The average negative cylinder magnitudes declined from -0.89 to 0.48 D (linear regression: p < 0.001). The J0 components displayed high correlations in both eyes (p < 0.001) but the J45 components did not. The average absolute anisometropias (difference of spheres) declined from 0.37 to 0.23 (linear regression: p < 0.001). Of the 736 children, 85 (11.5%) were referred to an eye care specialist. Of these, 52 received spectacles (61.2%), 14 (16.4%) were identified as "at risk" and remained under observation, and 18 (21.2%) were considered "false-positive."

Conclusions: Non cycloplegic photorefraction provides considerably less hyperopic readings than retinoscopy under cycloplegia. Additional refractions performed through binocular +3-D lenses did not facilitate detection of hyperopia. With the referral criteria above, 11% of the children were referred to an eye care specialist, but with a 20% false-positive rate. The screening had some power to identify children at risk but the number of false-negatives remained uncertain.

MeSH terms

  • Child, Preschool
  • Contraindications
  • Equipment Design
  • Germany / epidemiology
  • Humans
  • Infant
  • Mydriatics
  • Prevalence
  • Refraction, Ocular*
  • Refractive Errors / diagnosis*
  • Refractive Errors / epidemiology
  • Refractive Errors / physiopathology
  • Reproducibility of Results
  • Vision Screening / instrumentation*


  • Mydriatics