Defining real change in measures of stereoacuity.

PURPOSE To establish the thresholds for "real change" in stereoacuity by defining long-term test-retest variability as 95% limits of agreement for 4 stereoacuity tests. DESIGN Retrospective cohort study. PARTICIPANTS AND CONTROLS We identified 36 patients (median, 17 years; range, 7-76) with any type of stable strabismus who had stereoacuity measured on 2 consecutive visits. Stable strabismus was defined as angle of deviation within 5 prism diopters by simultaneous prism and cover test and prism and alternating cover test. METHODS Stereoacuity was measured at near using the preschool Randot and the near Frisby stereotests and at distance using the Frisby-Davis distance (FD2) and the distance Randot stereotests. Stereoacuity was transformed to log units for analysis. The 95% limits of agreement were calculated based on a 1.96 multiple of the standard deviation of differences between test and retest. MAIN OUTCOME MEASURES The 95% limits of agreement for change in stereoacuity thresholds at 2 consecutive visits. RESULTS The 95% limits of agreement were 0.59 log arcsec for the preschool Randot, 0.24 for the near Frisby, 0.68 for the FD2, and 0.46 for the distance Randot. These values correspond with the following octave steps (doublings of threshold; e.g., 200-400 arcsec): preschool Randot, 1.95; near Frisby, 0.78; FD2, 2.27; and distance Randot, 1.52. CONCLUSIONS A change of approximately 2 octaves of stereoacuity threshold are needed to exceed test-retest variability for most stereoacuity tests. Changes <2 octaves cannot be distinguished from test-retest variability. When used to guide patient management, caution should be taken in interpreting changes in stereoacuity of <2 octaves. FINANCIAL DISCLOSURE(S) The authors have no proprietary or commercial interest in any materials discussed in this article.