U.S. preference-weighted EQ-5D index scores were calculated on the basis of the method developed by Shaw et al.. All bleeds on all subjects were evaluated irrespective of treatment. Hierarchical models to control for repeated measures within patients, stratifi-cation, and comparison with mean nonbleed scores were used to evaluate instrument responsiveness to bleed and nonbleed days. QOL data were processed on all patients who had at least one bleeding episode and at least 7 days of QOL data past the end of the bleed. Additional post hoc analysis of this data included assessment of EQ-5D index, health, and pain scores across treatment groups and pediatric and adult age groups.