Overview of Methods Used in Biostatistics for the Analysis of Clinical Tiral Data

At first, it is essential to investigate the data’s basic character: distribution, outliers… If there is only a single measurement for each variable without covariate, the simple method of analysis may be an AVOVA or t-test or a non-parametric test (Wilcoxon). If the covariates belonging to the two comparison groups differ and are measured qualitatively an alnalysis of covariance (ANCOVA) is used. A balanced repeated measures ANOVA can be performed with of without covariates if the variable is measured more than once. Some studies result in a quantitative outcome variable and on or more quantitative covariates, multiple regression methods are useful for these kinds. For studies in which the outcome variable is qualitative, log-linear modeling Person Chi-square methods as the Cochran-Mantel-Haenzel method can be used. Logistic regression enables the statistician to determine relationships between the outcome variable and independent variables. The Cox regression method is adapted from survival analyses and can determine the effect of treatments and other potential covariates even when the data do not have the same follow-up time. Safety Analyses the number of patients with at least one adverse event can be compared the number of patients with at least one adverse event can be compared using Cochran-Mantel-Haenzel method or an equivalent procedure. International Conference on Harmonisation ICH, is concerned with harmonization of technical requirements for the registration of pharmaceutical products. Statistical Principles for Clinical Trials is discussed in E9 guideline. Conclusion: Internet Technology provides the possibility to perform the statistical programs respecting international norms everywhere in the world. The important recommendation is to use the analytical method that is most consistent with the data and best responds to the study objective and to international.

کلیدواژه ها:

clinical tiral ، biostatistician ، outliers ، ANOVA ، ANCOVA ، repeated measures ANOVA ، multiple regression ، loglinear modeling ، Cochran-Mantel-Haenzel ، Logistic regression ، Kaplan-Meier ، Cox regression ، Safety Analyses ، ICH