Repeatability of the phenol‐red thread and tear thinning time tests for tear film function

Abstract

Reliable tear function tests are required both in clinical practice and for research. Between‐sessions repeatability of the phenol‐red thread and tear thinning time tests was assessed for 17 young normal Caucasian subjects on two consecutive days. A single phenol‐red thread result and the mean of three tear thinning times were recorded on each day. The minimum phenol‐red thread result was 11 mm, and the 68 per cent and 95 per cent confidence intervals for repeatability were ± 5 mm and ± 10 mm, respectively. We propose diagnostic criteria of less than 11 mm for low tear secretion and less than 16 mm for a borderline result, based on the 68 per cent confidence interval. A level of measurement effect on repeatability of tear thinning time was present (Spearman's rho = 0.79, p = 0.002), with longer times showing greater variability: this relationship was eliminated by logarithmic data transformation. The tear thinning time test was not sufficiently reliable for clinical use; for a result of 15 seconds, there was 95 per cent confidence that subsequent times would lie between 57 and three seconds. We recommend further investigation into the use of log scales for tear stability tests.

Key words:

• tear function tests
• phenol‐red thread
• tear thinning time