②Syndrome was divided into 3 types: Yin deficiency with internal excessive heat (n=29), Qi-Yin deficiency (n=22) and deficiency of both Yin and Yang (n=30) according to the dialectic criterion concerning clinical research instruction in the treatment of diabetes mellitus of Traditional Chinese Medicine New Drug Clinical Instruction Principle designed by Department of Health in 1993 in combination of clinical practice.
Methods In conjunction with the instance of the clinical trial, to construct the correlation structures within subject, then to expatiate the theory about univariate ANOVA, Multivariate analysis of variance, mixed linear model and generalized estimating equation, and bring forth the step of these methods for new-drug clinical trial statistical analyse.
Results and Conclusions To show the applications of univariate ANOVA, Multivariate analysis of variance, mixed linear model and generalized estimating equation at new-drug clinical trial statistical analysis in conjunction with the instance. To demonstrate a series of resolve for non-independent data and to educe theirstrong suits, disadvantage and applicability.
A two-period cross-over design in new clinical trials is proposed. By the design the within-patient comparisons between drugs and between periods can be made. The tests of period-effect and drug-period interaction may explain whether the spontaneous changes of effect of periods and the carry-over effect of drugs exist respectively.