Abstract
PURPOSE: Detection of anisocoria in those with dark irides is difficult, and failure to detect anisocoria can have dire consequences. Whether infrared pupillometry and gross measurement would yield different prevalence rates for anisocoria in those with dark irides is unknown. We compared the frequency of anisocoria in healthy adults with dark irides assessed with mm ruler versus infrared pupillometry. METHODS: Pupil diameters in light (L) and dark (D) conditions were obtained to identify anisocoria in 59 human subjects with dark irides using two techniques. To avoid bias, gross measurements (S) with ruler were taken first. Pupils were imaged under infrared illumination mounted in a spectacle frame with mm tape attached. Adobe Photoshop was used to measure pupil sizes on the digital images (O). RESULTS: Proportions of anisocoria by group were SL .034, OL .130, SD 0.00, OD .135. Fisher's exact test showed that anisocoria in dim light was more frequent with the infrared photo technique. Exact binomial probability testing showed that the anisocoria in SL and SD conditions was not different from an expected proportion of 5%; whereas anisocoria in OL condition was not different from an expected proportion of 20%, and anisocoria in OD condition was not different from expected proportions of 10, 15, and 20%. CONCLUSIONS: In people with dark irides, ruler measurements of pupil size underestimate the frequency of anisoria in dim lighting conditions compared to the use of infrared pupillometry. Whether detection rates of pathologic anisocoria differ with measurement technique remain to be explored.