Purpose: To establish the strength of the association between routine tear function tests and conjunctival impression cytology (CIC) and to determine whether they simulate the morphological and cytological changes that occur on the ocular surface in dry eye. What are the sensitivity, specificity and positive predictive values of these tests when CIC is considered the gold standard?

Methods: The tear film profile included tear film break up time (TBUT), Schirmer's-1, Rose Bengal scores (RBS), and impression cytology. CIC samples were obtained from the inferior bulbar conjunctiva and stained with periodic acid-Schiff and counter stained with hematoxylin and eosin.

Results: The mean Schirmer's value was 11.66 ± 5.90 in patients and 17.17 ± 2.97 in controls (p < 0.001). The mean TBUT in participants was 8.88 ± 3.54 and 13.53 ± 2.12 in controls (p < 0.001). Patients had a mean goblet cell density (GCD) of 490 ± 213, while the value for controls was 1,462 ± 661 (p < 0.001). Abnormal CIC was observed in 46.7% cases of dry eye and in 32.8% of controls. The correlation coefficient (L) for Schirmer's was 0.2 and 0.24 for participants and controls, respectively, while TBUT values were 0.26 and 0.38, RBS were 0.5 and 0.5, and GCD was 0.8 and 0.6 in cases and controls, respectively.

Conclusions: GCD, RBS, and TBUT were better predictors of morphological and cytological changes in the conjunctiva than Schirmer's in dry eye syndrome. The sensitivity of tear function tests in diagnosing dry eye was TBUT > Schirmer's > RBS, and the specificity was Schirmer's > TBUT > RBS in decreasing order when CIC was considered the gold standard.