Dry Eye Disease: Concordance Between the Diagnostic Tests in African Eyes

Objective: To assess the concordance between the diagnostic tests for dry eye disease (DED) in a Nigerian hospital population. Methods: The study was a hospital-based cross-sectional survey of adults (>= 18 years) presenting at the eye clinic of the University of Nigeria Teaching Hospital (UNTH), Enugu; September-December, 2011. Participants' socio-demographic data were collected. Each subject was assessed for DED using the "Ocular Surface Disease Index" (OSDI) questionnaire, tear-film breakup time (TBUT), and Schirmer test. The intertest concordance was assessed using kappa statistic, correlation, and regression coefficients. Results: The participants (n = 402; men: 193) were aged 50.1 +/- 19.1 standard deviation years (range: 18-94 years). Dry eye disease was diagnosed in 203 by TBUT, 170 by Schirmer test, and 295 by OSDI; the concordance between the tests were OSDI versus TBUT (Kappa, kappa = 20.194); OSDI versus Schirmer (kappa = -0.276); and TBUT versus Schirmer (kappa = 0.082). Ocular Surface Disease Index was inversely correlated with Schirmer test (Spearman rho = -0.231, P < 0.001) and TBUT (rho = -0.237, P < 0.001). In the linear regression model, OSDI was poorly predicted by TBUT (beta = -0.09; 95% confidence interval (CI): -0.26 to -0.03, P = 0.14) and Schirmer test (beta = -0.35, 95% CI: -0.53 to -0.18, P = 0.18). Conclusion: At UNTH, there is poor agreement, and almost equal correlation, between the subjective and objective tests for DED. Therefore, the selection of diagnostic test for DED should be informed by cost-effectiveness and diagnostic resource availability, not diagnostic efficiency or utility.