PURPOSE. The purpose of this study was to compare two non-invasive methods for observing tear break-up, the Tearscope PlusTM (TS) and retroillumination of the tear film (RI), to the traditional fluorescein (FL) method. METHOD. Although the TS and RI (with an infrared light source) can be used on a non-dilated pupil, we dilated the eyes of 10 normal subjects with 1.0% tropicamide to enhance viewing of the TS image, and to quickly alternate among the three methods. After a period of 30 min, eyes were anesthetized with 0.5% proparacaine to allow subjects to keep their eyes open. While holding one eye closed, subjects were asked to keep the experimental eye open while the patterns of tear film disruption were videotaped alternately by the 3 techniques (TS, RI, and FL). Areas of tear break-up in digital images were circled by a masked observer, and the spatial distribution of tear breakup was compared by the three methods.
RESULTS. Tear break-up and tear film thinning was easily detectable by the traditional method of FL over any part of the cornea. RI detected tear film thickness differences with high resolution over the retroilluminated area of the pupil. RI appears to be an edge detector for tear film thickness differences, and thus does not differentiate between large areas of surface drying and an intact tear film. Areas of tear break-up were sometimes identifiable by TS, but tear break-up was often poorly visible, difficult to localize, or obscured by the lashes or other reflections. The spatial distribution of early tear break-up was highly correlated (r=0.79) between the RI and FL methods, but RI was less accurate when large areas of surface drying were present. Tear break-up visible by FL was frequently difficult to identify or was obscured in TS images.
CONCLUSIONS. RI produced an easily detected, high resolution image of tear break-up within the pupil that was highly correlated with the FL method unless large areas of surface drying were present. Tear break-up was more difficult to identify and localize using the TS, rendering this method less useful as a non-invasive method for measuring tear break-up.