5. Statistical analysis
According to literatures 7, 8, 12, 31-33, for the primary endpoint, we assumed that the 16-week CR rate was 25% in MA group, 50% in LNG-IUS group and 60% in MA+LNG-IUS group; with a power of 0.8 at a two-sided significance level of 0.05; requiring an accrual of 362 eligible patients (lost to follow-up rate <10%), which was too large to be carried out. Then we eventually decided to recruit 180 patients with 60 in each group as a phase II study. Modified intention-to-treat (ITT) analyses were performed for patients underwent endometrial evaluation at 16 or 32 weeks, and patients missed endometrial evaluation at 16 or 32 weeks but did not reach CR at subsequent endometrial evaluation. The latter was regarded as not reaching CR at 16 or 32 weeks. Patients missed endometrial evaluation at 16 or 32 weeks but reached CR at subsequent endometrial evaluation were excluded for 16 or 32-week CR rate analysis. ANOVA test or Kruskal-Wallis test was used for the comparison of continuous variables between the three groups, and Student’s t-test or Mann-Whitney test was used for comparison between two groups. Chi-square test or Fisher’s exact test were used for the differences in the categorical variable. Time-to-event endpoints were estimated with the Kaplan-Meier method. Log-rank test was used to compare the differences in survival curves. Cox regression analysis was used to estimate hazard ratio for CR or recurrence. A 2-tailed P -value of <0.05 was considered statistical significant. All statistical analyses were performed using SPSS for windows (version 22.0; Armonk, New York). COSORT guidelines were consulted to outline this study 34.