28 Feb Interrater reliability in pupillary measurement
Interrater reliability in pupillary measurement
Category: Critical Care
Witting MD, Goyal D., Interrater reliability in pupillary measurement., Ann Emerg Med. 2003 Jun;41(6):832-7.
STUDY OBJECTIVE: A previous report documented ranges of normal pupil size on the basis of measurements by the principal investigator. In this report, we examine interrater reliability of pupillary measurement. METHODS: According to a randomized double-blind assignment, healthy volunteers received phenylephrine in one eye and sterile water in the other. After a wait period, the principal investigator and then other observers, using a gauge with a modified Haab scale, took measurements of each pupil, both in fluorescent light (2,700 to 5,400 lux) and bright light (>54,000 lux), whereas the contralateral pupil was concealed. For the first pupil observed in each session, each observer also provided a gestalt judgment of whether the pupil was dilated. In the study’s main comparison, paired differences between principal investigator and nonprincipal investigator measurements were summarized by using the median and interquartile range for measurements taken in both light intensities. Interrater agreement in diagnosing pupillary dilation was also calculated for gestalt judgment and for simple measurement. RESULTS: Among 149 principal investigator/nonprincipal investigator pairs taken from 102 participants, principal investigator room-light measurements were 0.2 mm (median 0.2 mm; interquartile range -0.4 to 0.7 mm) greater than those from other observers, but there was no difference in bright-light measurements (median 0 mm; interquartile range -0.5 to 0.4 mm). Nevertheless, principal investigator/nonprincipal investigator percentage agreement in judging dilation by means of measurement was high in both light intensities (85% to 86%). In pupils with gestalt judgment of the presence or absence of pupillary dilation, percentage agreement between nonprincipal investigator observers was higher for measurement (75% to 82%) than for gestalt judgment (61%). CONCLUSION: There was no systematic difference between principal investigator and nonprincipal investigator bright-light measurements, supporting the range of normal values published previously. Interrater agreement in diagnosing pupillary dilation by means of measurement was high.