Abstract
Purpose :
To technically and clinically validate a new Liquid Jet (LJ) esthesiometer prototype for corneal sensitivity measurement, employing isotonic saline solution as stimulus at variable pressure levels and an examiner independent double-staircase method based algorithm for sensitivity threshold determination.
Methods :
i) Technical validation: Vertical and horizontal LJ displacement was recorded by means of a digital camera (Canon EOS 550D) at a distance of 15 mm to the exit nozzle four times at each setting, with a stimulus pressure range of 150 to 1250 mbar (stimulus duration: 40ms; stimulus temperature: 36°C).
ii) Clinical validation: Two corneal sensitivity thresholds were obtained with each, the Liquid Jet and Cochet Bonnet (CB) esthesiometer from 90 subjects on two separate days (4 measurements in total, gender balanced; approved by the Swiss Ethical Commission): Group A (18-30 years) and Group B (50-70 years). Inclusion criteria were no previous rigid or recent soft contact lens wear, and absence of ocular surface disease (OSDI </= 13.0).
Results :
i) Mean vertical (dv) and horizontal (dh) displacement decreased with increasing pressure: dv(150 mbar) = 0.67±0.09 mm and dv(1250 mbar) = 0.12±0.08 mm; dh(150 mbar) = 0.01±0.01 mm; dh(1250 mbar) = 0.11± 0.08mm.
ii) The Coefficients of Repeatability (CoR) for LJ measurements was 1.2 dB (signal to noise ratio (S/N): 3.7 dB; Bland Altman; data converted to a logarithmic scale) and for CB measurements, CoR was 1.2 dB and S/N was 3.8 dB. No significant difference in thresholds was observed between visits for the LJ method (p=0.86, Wilcoxon signed-rank test (WSRT)), however significantly higher thresholds were obtained at visit 2 with CB (p=0.027, WSRT). Significantly lower sensitivity was observed for LJ in Group B, however not for CB (LJ: p=0.035; CB: p=0.124, linear mixed effects model). Moderate correlation between LJ and CB was observed (r= 0.52, p<0.001, Spearman test).
Conclusions :
LJ centration on the ocular surface was good for the relevant stimulus range tested. Differences in corneal sensitivity thresholds between visits were only significant for the CB method, and an age dependency was observed for the LJ method. The thresholds obtained with the two different methods correlated moderately.
This is a 2020 ARVO Annual Meeting abstract.