Statistical methodology is an essential component of clinical (and biological) - search. Therefore it is not surprising that many textbooks aiming at ¿explaining¿ statistical methods to researchers have been published and continue to appear in print. The complexity of the basic issue, that of communication between st- isticians and researchers, is illustrated and discussed very well by the authors of this book in their Introduction. Thus it would be of no benefit to dwell further on this theme and on the difficulty of producing truly effective material. I prefer to start from a personal episode. Way back in 1959, a freshman of the faculty of Statistical Sciences, I found by chance on a stand, the book ¿Metodi statistici ad uso dei ricercatori¿, which was the Italian translation of the famous book by RA Fisher. The textbook for the first course of statistics used that year at of our University had left me perplexed. A student like myself, with a good high school scientific education, was fascinated by the wealth of real life ex- ples, but was unable to reconstruct the thread of logical-mathematical reas- ing, especially the inductive one. The unexpected access to Fisher¿s legendary book raised my hopes that I could finally get to the heart of the matter.

The scope of clinical research is to evaluate the effect of a treatment on the evolution of a disease in the human species. In recent years many introductory textbooks on clinical trial methodology have been published, some of which are excellent, in addition to a very extensive specialist literature. Here is a new book on methods and issues in clinical research. Its objectives can be summarized in three points. 1. Integrate medical and statistical components of clinical research. 2. Do justice to the operational and practical requirements of clinical research. 3. Give space to the ethical implications of methodological issues in clinical research. The book ends with a brief description of the drug development process and the phases of clinical development.

1. Variability of Biological Phenomena and Measurement Errors1.1. Phenotypic Variability1.2. Temporal Variability.1.3. Measurement - Related Variability1.3.1. The Measurement1.3.2. Measurement Errors1.4. Variability of Diagnostic TestsSummary2. Distinctive Aspects of a Biomedical Study. Observational and Experimental Studies2.1. Distinctive Features of Biomedical Studies2.2. The Study Protocol2.3. Observational Studies2.4. Experimental StudiesSummary3. Observational Studies3.1. Basic Designs of Observational Studies3.1.1. Prospective or Cohort Studies3.1.2. Retrospective Studies or Control Cases3.1.3. Sample Size3.2. Bias and Confounding3.2.1. Control of Bias in Epidemiology3.2.2. Control of the Phenomenon of Confounding |3.3. Advantages and Disadvantages of the Different Types of Observational StudiesSummary4 Defining the Treatment Effect4.1. From the Single Measurement to the Signal4.2. Identification and Quantification of the End-Points (Individual Subject Level)4.2.1. Methodological Characteristics of the End-Point4.2.2 Discriminating Between Primary and Secondary End-Points and Between Efficacy and Safety/Tolerability End-Points 4.3. Identification and Quantification of the Signals (Group Level)4.4. Statistical Considerations4.5. Practical, Regulatory, Marketing and Pharmaco-Economic Considerations4.6. Selection and Characterization of the Primary End-Point and Signal: an Example4.6.1. Stage One: Define the Main Therapeutic Level4.6.2. Stage Two: Define the Primary End-Point (Individual Patient Level)4.6.3. Stages Three and Four: Define the Group Indicator, the Signal, and the Threshold of Clinical Relevance (Treatment Group and Study Levels)4.7. More Than One Question in the Same Study: the Problem of Multiple Statistical Tests4.8. Validation of Measurement Scales4.9. Special Types of End-PointsSummary5. Probability, Inference and Decision Making5.1. Probability5.1.1 Definitions5.1.2. Probability Distribution and Probability Density Function5.1.3. Normal or Gaussian Distribution5.2. Basic Concepts of Inference5.2.1. Hypothesis Testing and Statistical Formulation of the Medical Question5.2.2. Statistical Estimation as the Tool for Evaluation of Clinical Relevance5.3. Statistical Inference in the Frequentist and the Bayesian Approaches5.4. Two Digressions: Measures of Variability and Likelihood Function5.4.1. Measures of Variability5.4.2. Likelihood Function5.5. Frequentist (Classical) Analysis of a Clinical Trial5.5.1. Hypothesis Testing: the Frequentist Solution5.5.2. Estimation of the Effect: the Frequentist Solution5.6. Bayesian Analysis of a Clinical Trial5.6.1. Hypothesis Testing: the Bayesian Solution5.6.2. Estimation of the Effect: the Bayesian Solution5.7. Some Additional Considerations on the Frequentist and Bayesian Approaches5.8. Parametric and Non-Parametric Inference5.9. Statistical Decision Making in the Medical Field5.10. Evidence-Based MedicineSummary6. The Choice of the Sample6.1. Which Subjects Should Form the Sample?6.1.1. Characteristics of the Patients to be Enrolled in the Study6.1.2. Mechanism of Subject Selection6.2. How Many Subjects Should Form the Sample?6.2.1. Statistical Considerations6.2.2. Medical and Practical AspectsSummary7. The Choice of Treatments7.1. Study Treatments7.1.1. How Many Treatments7.1.2. What Treatments7.1.3. Blinding of the Study Treatments7.1.4. Packaging and Logistics7.2. Concomitant TreatmentsSummary8. Experimental Design: Fallacy of 'Before-After' Compa