Figure 3.

Multiple Factorial Correspondence Analysis. Distance used: Chi-square. Axis 1 (QOL global score) by Axis 2 (Social Consequences). Active variables (the one that determined the structure and the identification of the axes) were the 25 items of the NEI-VFQ-25; they were not presented in this graph to improve readability. Illustrative variables (the one that allowed to identify patient sub-groups) consisted in side effect list of COM-TOL, compliance. Axis 1 estimated the intensity of QoL a global unique score: negative values were associated with 'Good QoL' while positive values were associated with 'Poor QoL'. Negative values on Axis 2 were associated with 'Bad social QoL'. Points, which were close, were highly correlated while points, which were far, were independent.