Agreement and repeatability of pupillometry using videokeratography and infrared devices

To evaluate the accuracy and repeatability of the pupil-measuring modules of several corneal topography devices.

Department of Ophthalmology, Saint Louis University Eye Institute, St. Louis, Missouri, USA.

In 14 eyes of 7 healthy myopic patients, pupillometry was performed with 3 corneal topography devices and with an infrared pupillometer under luminance-matched conditions for the Placido projection of the topography devices. Pupils were also measured under a mesopic condition. Outcomes were pupil diameters, limits of agreement, and coefficient of repeatability of the topography devices.

Mean pupil diameter measurements with the Technomed C-Scan, Humphrey Masterview, Alcon EyeMap, and under a mesopic condition were 3.35 mm, 2.96 mm, 2.34 mm, and 5.94 mm, respectively. All pupil diameter measurements differed significantly from one another except those by the Masterview and C-Scan devices. The mean difference between the C-Scan and luminance-matched infrared measurements was 0.74 mm and between the Masterview and luminance-matched infrared measurements, 0.27 mm. The limits of agreement +/- 2 standard deviations was 4.12 mm and 1.56 mm for the C-Scan and Masterview devices, respectively. Coefficients of repeatability were 0.56 mm, 0.46 mm, and 0.44 mm for the C-Scan, Masterview, and EyeMap devices, respectively.

Although topography pupillometry was repeatable, it underestimated the largest natural pupil diameter because of the luminance of the Placido rings. The difference in limits of agreement between the C-Scan and Masterview devices may be explained by pupil physiology and the static nature of videokeratoscopy. These results have implications when topography pupillometry is used to assess pupil diameters prior to refractive surgery. We do not recommend using pupil diameters measured by topography to preoperatively determine halo-related safety.