Efficacy and safety of Chou-Ling-Dan granules in the treatment of seasonal influenza via combining Western and traditional Chinese medicine: protocol for a multicentre, randomised controlled clinical trial

He, Jiayang; Li, Zhengtu; Huang, Wanyi; Guan, Wenda; Ma, Hongxia; Yang, Zi feng; Wang, Xinhua

SAP will be used to examine the laboratory data, adverse events and adverse reactions. The 399 rate of adverse reactions will employ the SAP as the denominator. i. SPSS17.0 or SAS9.2 statistical software will be used. 402 ii. All the statistical inferences will use two-sided tests. Statistically significant means P<0.05. The 403 confidence interval of the parameters will be estimated using a 95% confidence interval. 404 iii. Efficacy analyses of the LOCF will be used to compensate for the cases in which not all 405 information from the patients is recorded during treatment; the last observation data will transfer to 406 the final test results. The safety analysis will not estimate any missing data. 407 iv. A comparison between the dropout rate and dropout rate caused by adverse events will be 408 undertaken using the Pearson's χ 2 test. 409 v. Measurement data are given as mean, standard deviation and confidence intervals, and if 410 necessary, minimum, maximum, P25, median and P75 will be provided; paired measurement data 411 will also be used to demonstrate the difference between the mean and standard deviation; and with 412 non-parametric methods, the median and mean rank will be provided. Count data are the frequency 413 distribution and the corresponding percentage. Level data will be given as the frequency distribution 414 and the corresponding percentage, as well as the median and mean rank. Qualitative information 415 will show the positive rate, the number of positive cases and the denominator. 416 vi. Baseline data analyses (two sets) will involve demographic indicators, the general situation and 417 primary and secondary indicators before the intervention. Measurement data will be analysed using 418 the t test or t² test (when the variance is not homogeneous); count data will be analysed using the Pearson's χ 2 test. Rating data will be analysed using the two sample Wilcoxon rank sum test. 420 vii. For the efficacy analyses of quantitative variables, comparisons between groups will use 421 repetitive measure analysis of variance and covariance analysis. For qualitative variables, 422 comparisons between groups will use the Pearson's χ 2 test and centre effect analysis will use the 423 CMH test. For rating variables, group comparisons will be tested by the Kruskal-Wallis test and the 424 central effect analysis will be tested by the CMH test or grade logistic regression. 425 viii. For centre effect analysis, the CLM and CMH methods will be used for quantitative and 426 qualitative indicators, respectively. A logistic regression model will be performed for the evaluation 427 and correction of rating variables. 428 ix. With the subgroup analysis, it will be difficult to prevent the specific circumstances of the data 429 affecting the outcome. 430 x. For the safety analysis, the Pearson's χ 2 test will be used for comparing the prevalence of adverse 431 events (two sets) and the adverse events occurring in the trial will be listed and described. A 432 description of the laboratory test results of normal/abnormal changes before and after the 433 experiment, as well as the relationship between abnormal changes and the test drugs will be 434 recorded. These changes will also be listed and described. 435 Adverse events refer to any adverse medical events occurring from the time when patients 436 signed the informed consent form and were selected for the trial to the time of the last follow-up.