Statistics
Because interest was in those who gained at least the clinically important difference (MCID) of 0.1 HUI3 multi-attribute score units, change in HUI3 multi-attribute score was dichotomized at this value with a 1 representing achievement of at least the MCID and 0 otherwise. Fisher’s exact test was used to determine the association between the dichotomized HUI3 multi-attribute change score and variables of interest. Table 3 shows the variables considered with variables summarized using count and percent. Levels for the eight attributes were dichotomized into worse/better outcomes so that counts in each category were sufficiently large to detect significant differences between proportions, if they existed. Paired t-tests were used to examine change over time from baseline to year 1 follow-up. P-values less than 0.05 were considered significant. All analyses were conducted in R4.