Table 2.

Results of efficacy analysis

Least square means (LSM) of slope and p-values were obtained from the analysis of covariance model with change from baseline (slope) as the dependent variable, treatment (Oral solubilized UDCA vs placebo) as factor, and the period and sequence as covariates. Values in paresthesis are standard errors of LSM. Efficacy was analyzed separately for each outcome without imputation for missing data, and definitions of the PP and PP' datasets were as followings: PP datasets, the data from the patients (n) who completed either the first or second period of study; PP' datasets, the data from the patients (N) who completed both the first and second periods of study.