In 35 patients with mild, moderate and severe keratoconjunctivitis sicca, an association was found between treatment effect, break-up time value and Rose bengal score. Neither of these tests, used separately, was successful as a clinically valuable predictor of treatment effect as the scatter, in score points, was too large. Analysed together, however, as in a partial regression analysis, an adequate prediction is possible, which is clinically of some value.
展开▼