Primary Efficacy Analysis. Response efficacy endpoints from Week 1 through Week 12 will be analyzed using the Xxxxxxx -Xxxxxx-Haenszel risk difference estimate. The estimated difference in the proportion of responders and 90% confidence interval for the difference will be calculated using the Xxxxxxx -Xxxxxx-Haenszel risk difference estimate stratified by baseline abdominal pain (< 6 vs ≥ 6) with weights proposed by Greenland and Robins. For the analysis of other continuous endpoints (eg, change from baseline at Week 12), a mixed model for repeated measures with restricted maximum likelihood estimation will be used by incorporating on treatment- values at all time points. The analysis model for the efficacy endpoint will include terms for treatment, visit, baseline score, and interaction of visit by treatment. An unstructured covariance matrix will be used to model the correlation among repeated measurements. The Xxxxxxx-Xxxxx adjustment for denominator degrees of freedom will be used with restricted (or residual) maximum likelihood to make statistical inference. Further details will be provided in the SAP.
Primary Efficacy Analysis. The primary efficacy endpoint is the responder rate for lip fullness on the LFS2, which is defined as the proportion of subjects who show ≥ 1-point improvement on the LFS2 compared to baseline assessment, at Day 30 after last treatment received.
Primary Efficacy Analysis. The efficacy evaluation will be based on the full analysis set, which will include all patients who receive at least 1 dose of study drug. The statistical method to be used for the SRI is the logistic regression model with treatment and stratification factors as main factors. Patients who withdraw from the study are classified as nonresponders as measured by SRI at week 52 in the primary analysis. The SRI is linked to the model factors through the logit function. The likelihood-ratio–based Chi-square statistics will be used for testing the treatment difference between IPP-201101 and placebo in SRI at week 52 at the significance level 0.05. The odds ratio (active/placebo) and associated 95% confidence interval will be determined from the logistic regression model. As sensitivity analyses, the primary analysis will be repeated using another 3 imputation methods to estimate missing SRI at week 52 according to the reason for withdrawal. In the 1st method, patients who withdraw because of lack of efficacy will be classified as treatment failures assessed by SRI at week 52. For patients who withdraw because of other reasons, their missing SRI at week 52 will be estimated using the multiple imputation method (Xxxxxx and Xxxxx, 2002). The imputation values will be drawn from the categories formed by treatment group and randomization stratum to which the patient belongs. In the 2nd method, the missing SRI at week 52 will be imputed using the last observation carried forward (LOCF) method. In the 3rd method, patients who withdraw and those completers who used prohibited medication within 8 weeks from week 52 will be classified as nonresponders.
