A Randomized, Single-blind, Group sequential, Active-controlled Study to evaluate the clinical efficacy and safety of a-Lipoic acid for critically ill patients with coronavirus disease 2019(COVID-19)

Ming Zhong; Aijun Sun; Ting Xiao; Ge Yao; Ling Sang; Xia Zheng; Jinyan Zhang; Xuejuan Jin; Lei Xu; Wenlong Yang; Peng Wang; Kai Hu; Dingyu Zhang; Junbo Ge

We estimated that with 10 deaths the study would have 68% power to detect a 50% improvement in survival with ALA versus standard control group (from admission to discharge), at a two-sided alpha level of 0.05. Primary and secondary endpoints were measured in the intention-to-treat population, which included all patients who underwent randomization. The Kaplan-Meier method was used to estimate time-to-event end points. Two-sided stratified log-rank tests and 2Log (LR) test were used. Cox regression models were applied to estimate hazard ratios. Safety evaluations were included all patients who received at least one dose of study medication. Continuous variables were described using mean, median, and standard deviation, and examined using Analysis of Variance (ANOVA) or Wilcoxon rank sum test or Kruskal-Wallis test. Categorical variables were described by frequency, and assessed using chi-square test or Fisher exact test. Changes in laboratory data, signs and examination items, and adverse events before and after treatment in the two groups will be analyzed as safety indicators. This study does not estimate missing data. Statistical analysis was performed using SAS 9.13. Values of P < .05 were considered statistically significant, and the two-sided test is represented by a 95% confidence interval.