Efficacy and utility of clinical examination in predicting meibomian gland atrophy.

PURPOSE: To investigate the role of clinical examination findings in predicting meibomian gland atrophy (MGA).

METHODS

A single-center, cross-sectional study was conducted. Subjective reports of dry eye symptoms were collected via a SPEED questionnaire. Lipid layer thickness, Schirmer I score, lid margin characteristics (thickening, vascular engorgement, and telangiectasia), and meibomian gland secretion factors (quality, expressibility, and volume) were examined and used to generate Foulks-Bron scores. Infrared meibography determined the degree of MGA. Multivariate binary logistic regression analysis determined predictive factors.

RESULTS

When all clinical characteristics were included, the pseudo R2 of the regression model was 0.466 (p-value = 0.029). FB score alone was significantly correlated with MGA severity (R = 0.372, p = 0.014) and had an odds ratio of 1.190 (p = 0.012). Expressibility was the main predictor for severity of MGA with an odds ratio of 2.604 (p = 0.020), sensitivity of 85.7 %, and specificity of 59.1 %. Lipid layer thickness (p = 0.042) and vascular engorgement (p = 0.032) were also predictive of MGA.

CONCLUSION

A thorough physical examination emphasizing manual expression of meibomian glands can be useful for predicting gland atrophy. Nevertheless, most of the variability in MGA cannot be explained by clinical characteristics, and the full spectrum of contributing factors remains incompletely understood.