Fig. 1

Bland-Altman Plot of actual, patient-reported EQ-5D-3L preference weights versus EQ-5D-3L preference weights estimated from SF-12 responses. Each marker represents one patient actual-estimated pair. The x-axis shows the mean of the patient-reported, actual, EQ-5D-3L preference weights, and the EQ-5D-3L preference weights estimated from SF-12 responses. The y-axis shows the difference between the actual and estimated preference weights. The solid line represents the overall mean difference between actual and estimated preference weights scores, and the dashed lines represent the limits of agreement (1.96 SD mean difference), which include 95 % of differences between actual and estimated responses. Where perfect agreement is observed, individual points line up along the 0 line of the y-axis