Comparison of 2 multiple-measurement infrared pupillometers to determine scotopic pupil diameter

To compare a monocular and a binocular multiple-measurement digital infrared pupillometers for measuring scotopic pupil diameter.

Department of Ophthalmology, Loma Linda University, Loma Linda, California, USA.

Scotopic pupil size was measured after 1 minute of dark adaptation in 42 eyes of 21 volunteers. Measurements were taken twice each with 2 multiple-measurement digital infrared pupillometers, the monocular pupillometer (Neuroptics, Inc.), and the binocular pupillometer (P2000D, Procyon, Ltd.) Intraclass correlation coefficients (ICC) and limits of agreement (LOA) were used to measure repeatability and agreement of measures with each instrument and between instruments. The Wilcoxon signed rank test was used to compare variability of measurements within each instrument.

The mean scotopic pupil size was 4.79 mm +/- 0.95 (SD) with the Procyon and 4.86 +/- 0.93 mm with the Neuroptics. Repeatability and agreement tests for the Procyon measures showed the following: ICC, 0.954; 95% confidence interval (CI), 0.916-0.975; LOA, -0.60 to 0.56; range, 1.16. The Wilcoxon signed rank test of variability gave a Z score of -2.53 (P = .01, 2 tailed). The repeated measures testing with the Neuroptics pupillometer showed the following: ICC, 0.985; 95% CI, 0.972-0.992; LOA, -0.39 to 0.26; range, 0.64; Z score, -1.15 (P = .25, 2-tailed). Repeatability and agreement tests for measures between instruments showed the following: ICC, 0.954; 95% CI, 0.916-0.975; LOA, -0.60 to 0.50; range 1.11.

There was a high repeatability and agreement in scotopic pupil diameter for repeated measures within each device and measurements between the devices. Differences in variability in scotopic pupil diameter evaluated by the Wilcoxon signed rank test were significant only with the Procyon pupillometer.