ABSTRACT: Purpose. To investigate within‐ and between‐examiner reliability and validity of hand‐held refractometry water content measures of hydrogel lenses. Methods. Nineteen lenses of various nominal water contents were examined by two examiners on two occasions separated by 1 hour. An Atago N2 hand‐held refractometer was used for all water content measures. Lenses were presented in a random order to each examiner by a third party, and examiners were masked to any potential lens identifiers. Intraclass correlation coefficients (ICC), 95% limits of agreement, and Wilcoxon signed rank test were used to characterize the within‐ and between‐examiner reliability and validity of lens water content measures. Results. Within‐examiner reliability was excellent (ICC, 0.97; 95% limits of agreement, −3.6% to +5.7%), and the inter‐visit mean difference of 1.1 ± 2.4% was not biased (p = 0.08). Between‐examiner reliability was also excellent (ICC, 0.98; 95% limits of agreement, −4.1% to +3.9%). The mean difference between examiners was −0.1 ± 2.1% (p = 0.83). The mean difference between the nominally reported water content and our water content measures was −2.1 ± 1.7% (p < 0.001); the 95% limits of agreement for this difference were −5.4% to +1.1%. Conclusion. There is good reliability within and between examiners in measuring water content of hydrogel lenses. However, with our sample of lenses, examiners tended to overestimate the nominal water content of hydrogel lenses. As discussed, this bias may be associated with the Brix scale used in refractometry and is material dependent. Therefore, investigators may need to account for bias when measuring hydrogel lens water content via hand‐held refractometry.
Nichols et al. (Sun,) studied this question.