Measurement variability of the TearLab Osmolarity System.

Purpose: To independently assess the measurement variability of TearLab System in a clinical setting of one visit and to estimate the minimum number of measurements required for reliable readings of tear osmolarity.

Methods: Ten consecutive osmolarity measurements were taken from both eyes by the same examiner at one visit for fourteen subjects. The ocular surface disease index symptoms questionnaire and tear film break up time were also performed. Group average cumulative mean and cumulative coefficient of variation were calculated to assess the TearLab measurement variation. Repeated application of Thompson's tau method was performed to identify the outliers in tear osmolarity readings for each eye. Results from both eyes were analysed separately.

Results: Up to two randomly occurring outlying values in 10 consecutive measurements were found in 19 out of 28 measured eyes. No statistically significant differences between the left and right eye were found for the group mean and group standard deviation (paired t-test, p=0.099 and p=0.068, respectively), however the cumulative coefficient of variation indicated higher measurement group variability on one eye. Estimated cumulative coefficient of variation indicated the minimum of three consecutive acquisitions required for the measurement to be reliable.

Conclusions: TearLab Osmolarity System required at least three consecutive measurements to be taken in order to provide clinically reliable tear osmolarity readings. Also, taking the maximum osmolarity value for detecting dry eye disease should be viewed with caution since outlying readings of tear osmolarity frequently occur.