Lack of Agreement among Electrical Impedance and Freezing-Point Osmometers

Purpose. To assess the interchangeability of tear osmolarity measurements between electrical impedance and freezing-point depression osmometers and to analyze inter-eye tear osmolarity variability measured with these osmometers in healthy subjects. Methods. Tear osmolarity was measured using the TearLab osmometer (OcuSense Inc., San Diego, CA) and the Fiske 210 microsample osmometer (Advanced Instruments Inc., Norwood, MA). We randomly selected one eye in 50 subjects (29 women, 21 men; mean age, 33.16 +/- 6.11 years) to analyze whether osmolarity measurements by these osmometers were interchangeable. Both eyes of 25 patients (15 women, 10 men; mean age, 34.32 +/- 6.37 years) were included to analyze inter-eye osmolarity variability. Results. The mean tear osmolarity values measured with the TearLab osmometer were higher (305.22 +/- 16.06 mOsm/L) than those with the Fiske 210 osmometer (293.40 +/- 12.22 mOsm/L), with the intraclass correlation coefficient being 0.23 (p = 0.051). A Bland-Altman plot showed that the systems were not interchangeable because there was a systematic difference, with the limits of agreement being -17.93 to 41.57 mOsm/L. There were no statistically significant differences (p = 0.5006 and p = 0.6533, respectively) between an individual's eyes measured with either osmometer. Conclusions. Because the TearLab tear osmolarity measurements were higher than those of the Fiske 210 measurements and the limits of agreement were too wide, the two osmolarity values cannot be used interchangeably. In healthy subjects, there is no difference in tear osmolarity between right and left eyes of the same individual measured with both instruments.