Reproducibility of Liquid Jet Aesthesiometer to Measure Corneal Sensitivity

Abstract

Purpose : To determine the clinical reproducibility of corneal sensitivity threshold measurements using the newly developed liquid jet aesthesiometer (LJA).

Methods : The LJA (Brien Holden Vision Institute, Sydney) propels tiny droplets of sterile phosphate buffered saline (PBS) onto the central cornea under controlled conditions to generate a mechanical stimulus of variable strength. Sensitivity threshold is determined using an automated double staircase method and subjective response via a handheld pushbutton. 30 subjects, 58% female, 67% regular soft contact lens (SCL) wearers, aged 20 to 50 years completed the study. Each subject was measured on 5 different days at the same time of the day by one of six randomly assigned operators. Stimulus strength was varied by the volume of PBS projected onto the cornea (0.15 to 4.75 µl) with each stimulation. Stimulation duration was 160 ms, PBS warmed up to 36°C. Threshold values were taken as the average of the last 9 stimuli volumes after the ‘start high’ and ‘start low’ staircases have converged for the first time. A maximum of 40 stimuli were presented for each measurement. Coefficient of reproducibility was computed as 1.96 x within subject SD from ANOVA.

Results : The mean sensation threshold for all subjects and all visits was 2.4 µl, ranging from 0.19 to 4.66 µl for all patients and visits. The standard deviation between the 5 visits ranged from 0.14 to 1.79 µl, with the level of SD showing no correlation with the average threshold value R² = 0.0018. The SD within subjects was 0.85 µl, leading to a coefficient of reproducibility of 1.66 µl. Excluding the first and last visit, the respective values improved marginally to 0.77 and 1.50 µl. Sensation threshold and reproducibility were not significantly different between SCL and non-SCL wearers. In 6 out of the 150 measurements, the double staircase did not converge to a reliable endpoint.

Conclusions : Within a single visit, the staircase converged to a stable threshold endpoint for more than 96% of measurements, indicating a consistent stimulus and consistent subjective responses. No learning effect was observed, a single measurement should be sufficient to obtain a valid sensitivity values. However, for half the subjects, the day-to-day reproducibility was too high to be of clinical use. Further investigations will determine how reproducibility can be improved by fine-tuning the instrument or procedures.

This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.

 

SD in µl for visits 2-4