Purpose: To determine the repeatability of TearLab and I-PEN osmometers in vivo and their accuracy in vitro.
Design: Prospective, single-visit study.
Methods: The tear osmolarity of 28 participants was evaluated with TearLab and I-PEN on two occasions in random order, over a 2-h period. Both eyes were measured in a randomised order. Coefficients of repeatability (CoR) were determined for each device, together with the bias and limits of agreement between them. For the in vitro experiment, the osmolarity was measured by both osmometers in five solutions (290, 297, 342, 338 and 383 mOsm/L) at two different temperatures (22 and 37°C) with a total of four consecutive measures.
Results: The CoRs for the TearLab and I-PEN in the right and left eyes were 26.2, 21.3, 33.6 and 28.3 mOsm/L, respectively. Across the first and second repeats, TearLab showed consistency of diagnosis for 50% of participants with 29% as dry eye positive, while I-PEN indicated 68% consistency of diagnosis with 57% dry eye positive. The instruments agreed on the diagnosis in 46.5% of cases. In vitro comparison showed that the average measurement errors for TearLab and I-PEN were -10 ± 13 and 31 ± 39 mOsm/L at 22°C, and 4 ± 13 and 20 ± 51 mOsm/L at 37°C.
Conclusions: In vitro, both instruments showed reasonable accuracy and repeatability at mid-range osmolarities, but repeatability generally declined at higher and lower levels. While TearLab accuracy remained consistent across the osmolarity range, measurement errors for I-PEN noticeably increased outside the mid-range. In vivo, both instruments displayed poor repeatability. This casts doubt on the value of utilising either instrument to establish osmolarity as a factor in the diagnosis of dry-eye, according to currently recommended diagnostic guidelines (TFOS DEWS II), if only a single measurement is taken from each eye.
Keywords: I-PEN; TearLab; diagnostic performance; dry eye disease; tear osmolarity.
© 2021 The Authors Ophthalmic and Physiological Optics © 2021 The College of Optometrists.