PURPOSE: To determine the difference in pupil size measured with the Colvard pupillometer in mesopic and scotopic luminance and with the Rosenbaum pupil card in mesopic luminance between 2 examiners. SETTING: Michel Pop Clinics, Montreal, Quebec, Canada. METHODS: Two examiners used the Colvard pupillometer and the Rosenbaum card to measure pupil size in 58 eyes. The Colvard pupillometer was used in mesopic and scotopic light conditions. The Rosenbaum card was used in mesopic luminance only. Pupil size was evaluated with a 1.0 mm interval scale at the nearest half millimeter. RESULTS: For the 3 sets of data, the limits of agreement and coefficient of interrater repeatability were calculated and a 2 x 2 factorial analysis of variance was performed. Because of interexaminer bias, measurements done in mesopic luminance with the Rosenbaum card were not statistically different from those with the Colvard pupillometer in scotopic luminance, although interrater repeatability of the Colvard pupillometer (0.8 mm) was superior to that of the Rosenbaum card (1.3 mm). CONCLUSIONS: Examiner bias was the greatest statistical bias in all sets of measures. Surgeons may want to opt for a "safe" limit of pupil size (ie, 0.5 to 0.8 mm greater than the measured size) when calculating optical zones in refractive surgery. Future devices for pupil measurement should be based on automatic adjustment sizing.
PURPOSE: To determine the difference in pupil size measured with the Colvard pupillometer in mesopic and scotopic luminance and with the Rosenbaum pupil card in mesopic luminance between 2 examiners. SETTING: Michel Pop Clinics, Montreal, Quebec, Canada. METHODS: Two examiners used the Colvard pupillometer and the Rosenbaum card to measure pupil size in 58 eyes. The Colvard pupillometer was used in mesopic and scotopic light conditions. The Rosenbaum card was used in mesopic luminance only. Pupil size was evaluated with a 1.0 mm interval scale at the nearest half millimeter. RESULTS: For the 3 sets of data, the limits of agreement and coefficient of interrater repeatability were calculated and a 2 x 2 factorial analysis of variance was performed. Because of interexaminer bias, measurements done in mesopic luminance with the Rosenbaum card were not statistically different from those with the Colvard pupillometer in scotopic luminance, although interrater repeatability of the Colvard pupillometer (0.8 mm) was superior to that of the Rosenbaum card (1.3 mm). CONCLUSIONS: Examiner bias was the greatest statistical bias in all sets of measures. Surgeons may want to opt for a "safe" limit of pupil size (ie, 0.5 to 0.8 mm greater than the measured size) when calculating optical zones in refractive surgery. Future devices for pupil measurement should be based on automatic adjustment sizing.
Authors: Miguel J Maldonado; Alberto López-Miguel; David P Piñero; José R Juberías; Juan C Nieto; Jorge L Alió Journal: Graefes Arch Clin Exp Ophthalmol Date: 2012-01-08 Impact factor: 3.117