Pupillometer Inter-Device Reliability Assessment

INTRODUCTION: Assessment of the pupillary reflex is one of the fundamental aspects of the neurological examination. Traditionally, it comprises of a subjective assessment of the size and shape of the pupil prior to the manual application of a light source and the speed of pupil reactivity when exposed to light. We have recently shown that there is a striking degree of disagreement between trained observers in interpreting pupillary size and reactivity, partly because examiners are allowed to use a variety of non-standardized handheld light sources. Automated hand-held pupillometers are commercially available that are able to provide an objective measurement of the initial and final size of the pupil and to grade the speed of pupil contraction in response to a light stimulus. The purpose of this prospective study was to assess the inter-device reliability of the pupillary examination performed with NPiTM-100 pupillometers (NeuroOptics, Inc.). METHODS: To determine the inter-device reliability, 33 practitioners (28 RNs, 2 MDs, 1 NP, 1 medical student and 1 PhD research coordinator) obtained 210 paired pupillometer measurements from 20 patients at risk for cerebral edema. Paired pupillometer assessments were completed within a five-minute period and consisted of two separate assessments by two different clinicians using two different pupillometers. RESULTS: There was no statistically significant difference between the mean maximum pupil size at rest for both OS (p=0.27) and OD assessments (p=0.74) when measured by the different pupillometers (PM1 and PM2) prior to the light stimulation. Similarly, there was no statistically significant difference between the mean minimum pupil size for both OS (p=0.64) and OD (0.44) when measured after the light stimulation. The mean pupil reactivity for both OS (p=0.36) and OD (p=0.82) are also not statistically significantly different. In addition, Cohen's Kappa assessments of pupil size and reactivity revealed an almost perfect agreement between PM1 and PM2 for both the maximum pupil size of both OS (k=0.97) and OD observations (k=0.91) and the minimum pupil size of OS (k=0.96) and OD observations (k=0.98). Including the non-reactive pupil results, there was also a high correlation for the OS (k=0.99) and OD readings (k=0.90). DISCUSSION: The NPiTM-100 pupillometers have a high inter-device reliability, even when two different pupillometer devices are used on the same subject by different practioners. The data provide sufficient support to conclude that the use of NPiTM-100 pupillometers is likely to improve the precision of the pupillary examination.