Evaluation of the glasses-free Bernell evaluation of stereoacuity test

The Bernell Evaluation of Stereopsis Test (BEST) uses lenticular technology to present engaging, full-color images while minimizing monocular cues. We compared the performance of BEST with the Randot Preschool Stereoacuity Test (RPST) and Randot test (RT) in a pediatric cohort.

Children with strabismus, anisometropia, cataracts, and control children were included to provide diversity of stereoacuity scores in the cohort (n = 108; 42% female). In alternating order, children attempted BEST dinosaur (gross stereo screening) and graded disparity clip art tests as well as the RPST and RT tests. A subset (n = 25) who had no intervening treatment repeated BEST 1-6 months later. Stereoacuity thresholds were converted to log arcsec and tests were compared using the Fisher exact test and Bland Altman analysis.

For children 2-3.9 years of age (n = 25), the completion rate was 64%, 64%, and 67% for BEST, RPST, and RT, respectively. We were unable to identify a significant association between results from BEST and RPST (P = 0.46) or BEST and RT (P = 0.99). For children 4-7 years of age (n = 83), the completion rate was 94%, 100%, and 100% for BEST, RPST, and RT, respectively. There was a significant association between results from BEST and RPST (P < 0.001) and BEST and RT (P < 0.0001) results. Among 2- to 7-year-olds, BEST test retest reliability was poor, with limits of agreement of -0.80 to 0.83 log arcsecs.

BEST showed a high completion rate and strong associations with established stereoacuity tests for children 4-7 years of age but not for those 2-3.9 years of age. Overall, test-retest reliability was poor.