Purchase this article with an account.
Trefford L Simpson, Nancy J Keir, Yunwei Feng, William Ngo; Using detection thresholds to select ocular surface sensory panels II: Criteria based on mechanical and chemical threshold combinations.. Invest. Ophthalmol. Vis. Sci. 2017;58(8):3054. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To determine if we can we select ‘good’ and ‘bad’ participants in a sensory panel who were able to interocularly judge contact lens comfort and match it to pneumatic stimuli delivered using a Belmonte esthesiometer.
49 adapted contact lens wearers had mechanical and chemical detection thresholds (flow rate [ml/min] and % added CO2, respectively), measured with the Waterloo Belmonte Esthesiometer using the ascending method of limits. For each stimulus, thresholds were estimated on 3 occasions and averaged. From the average mechanical and chemical thresholds, subjects were classified as being sensitive (both thresholds below the median, n=17), insensitive (both thresholds above the median, n=16) or mixed (n=16). Subjects also interocularly matched the discomfort of 3 optimally fit contact lenses (each on a randomly selected eye) using the mechanical or chemical stimuli of the esthesiometer (obviously stimulating the other eye) and scaled contact lens discomfort using magnitude estimation. We tested a simple hypothesis: Subjects in the ‘good’ group would be better at interocular matching and the regression of the pneumatic stimulus match to rated discomfort would be linear with a positive slope, but the other groups would not perform well, with random slopes (positive, 0 and negative). Bayesian mixed modeling (with vague priors) estimated the group slopes (among others) and the distributions of these estimates were compared in the sensitive and other groups defined by detection thresholds. R and rjags were used for the data analysis.
For chemical matching, the sensitive group's 95% HDI of the Bayesian estimate of the interocular-match regression slope was 0.25-0.86. In the insensitive and mixed groups, the 95% HDI’s both included zero. For mechanical matching, the sensitive group's slopes were the highest but the 95% HDI included zero. The figure shows the posterior distribution of the slope estimates in the sensitive group.
Pneumatic thresholds can be used to separate sensitive and insensitive subjects who can match contact lens discomfort in a ‘well behaved’ way and therefore pneumatic esthesiometry may be used to select sensory panels. A simple rule to select these subjects is an exclusionary one: If subjects’ mechanical or chemical thresholds are above the median, exclude them – they should not be part of the sensory panel.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.
This PDF is available to Subscribers Only