Full Description
Statistical concepts provide scientific framework in experimental studies, including randomized controlled trials. In order to design, monitor, analyze and draw conclusions scientifically from such clinical trials, clinical investigators and statisticians should have a firm grasp of the requisite statistical concepts. The Handbook of Statistical Methods for Randomized Controlled Trials presents these statistical concepts in a logical sequence from beginning to end and can be used as a textbook in a course or as a reference on statistical methods for randomized controlled trials.
Part I provides a brief historical background on modern randomized controlled trials and introduces statistical concepts central to planning, monitoring and analysis of randomized controlled trials. Part II describes statistical methods for analysis of different types of outcomes and the associated statistical distributions used in testing the statistical hypotheses regarding the clinical questions. Part III describes some of the most used experimental designs for randomized controlled trials including the sample size estimation necessary in planning. Part IV describe statistical methods used in interim analysis for monitoring of efficacy and safety data. Part V describe important issues in statistical analyses such as multiple testing, subgroup analysis, competing risks and joint models for longitudinal markers and clinical outcomes. Part VI addresses selected miscellaneous topics in design and analysis including multiple assignment randomization trials, analysis of safety outcomes, non-inferiority trials, incorporating historical data, and validation of surrogate outcomes.
Contents
Part I. Introduction to Randomized, Controlled Trials. 1. Introduction. Part II. Analytic Methods for Randomized, Controlled Trials. 2. Dichotomous and ordinal: chi-square and Fisher's exact tests and binary regression models. 3. Continuous: t-test, Wilcoxon-test, and linear or non-linear regression models. 4. Time to event subject to censoring: logrank test, Kaplan-Meier estimation and Cox proportional hazards regression models. 5. Count: Poisson and negative binomial regression models. 6. Longitudinal: Linear and generalized linear mixed models, GEE. 7. Recurrent events. 8. Cross-over design. 9. Factorial design. 10. Cluster randomized design. 11. Randomization, stratification, and outcome-adaptive allocation. 12. Sample size estimation and power analysis: Dichotomous, ordinal, continuous and count. 13. Sample size estimation and power analysis: Time-to-event data subject to censoring. 14. Sample size estimation and power analysis: Longitudinal data. 15. Group sequential methods, triangular methods and stochastic curtailments. 16. Sample size re-estimation. 17. Adaptive designs. 18. Multiple testing. 19. Subgroup analysis. 20. Competing risks. 21. Joint models for longitudinal markers and clinical outcomes. 22. Sequential multiple assignment randomization trial (SMART) for dynamic treatment allocation. 23. Safety data analysis. 24. Non-inferiority trials. 25. Incorporating historical data into RCTs. 26. Validation of surrogate outcomes.
