Clinical Epidemiology and Biostatistics: A Primer for Clinical Investigators and Decision-Makers

ISBN-13: 9783642648144 / Angielski / Miękka / 2011 / 286 str.

Michael S. Kramer

Opis

Here is a book for clinicians, clinical investigators, trainees, and graduates who wish to develop their proficiency in the planning, execution, and interpretation of clinical and epidemiological research. Emphasis is placed on the design and analysis of research studies involving human subjects where the primary interest concerns principles of analytic (cause-and- effect) inference. The topic is presented from the standpoint of the clinician and assumes no previous knowledge of epidemiology, research design or statistics. Extensive use is made of illustrative examples from a variety of clinical specialties and subspecialties. The book is divided into three parts. Part I deals with epidemiological research design and analytic inference, including such issues as measurement, rates, analytic bias, and the main forms of observational and experimental epidemiological studies. Part II presents the principles and applications of biostatistics, with emphasis on statistical inference. Part III comprises four chapters covering such topics as diagnostic tests, decision analysis, survival (life-table) analysis, and causality.

Spis treści

I Epidemiologic Research Design.- 1: Introduction.- 1.1 The Compatibility of the Clinical and Epidemiologic Approaches.- 1.2 Clinical Epidemiology: Main Areas of Interest.- 1.3 Historical Roots.- 1.4 Current and Future Relevance: Controversial Questions and Unproven Hypotheses.- 2: Measurement.- 2.1 Types of Variables and Measurement Scales.- 2.2 Sources of Variation in a Measurement.- 2.3 Properties of Measurement.- 2.4 “Hard” vs “Soft” Data.- 2.5 Consequences of Erroneous Measurement.- 2.6 Sources of Data.- 3: Rates.- 3.1 What is a Rate?.- 3.2 Prevalence and Incidence Rates.- 3.3 Stratification and Adjustment of Rates.- 3.4 Concluding Remarks.- 4: Epidemiologic Research Design: an Overview.- 4.1 The Research Objective: Descriptive vs Analytic Studies.- 4.2 Exposure and Outcome.- 4.3 The Three Axes of Epidemiologic Research Design.- 4.4 Concluding Remarks.- 5: Analytic Bias.- 5.1 Validity and Reproducibility of Exposure-Outcome Associations.- 5.2 Internal and External Validity.- 5.3 Sample Distortion Bias.- 5.4 Information Bias.- 5.5 Confounding Bias.- 5.6 Reverse Causality (“Cart-vs-Horse”) Bias.- 5.7 Concluding Remarks.- 6: Observational Cohort Studies.- 6.1 Research Design Components.- 6.2 Analysis of Results.- 6.3 Bias Assessment and Control.- 6.4 Effect Modification and Synergism.- 6.5 Advantages and Disadvantages of Cohort Studies.- 7: Clinical Trials.- 7.1 Research Design Components.- 7.2 Assignment of Exposure (Treatment).- 7.3 Blinding in Clinical Trials.- 7.4 Analysis of Results.- 7.5 Interpretation of Results.- 7.6 Ethical Considerations.- 7.7 Advantages and Disadvantages of Clinical Trials.- 8: Case-Control Studies.- 8.1 Introduction.- 8.2 Research Design Components.- 8.3 Analysis of Results.- 8.4 Bias Assessment and Control.- 8.5 Advantages and Disadvantages of Case-Control Studies.- 9: Cross-Sectional Studies.- 9.1 Introduction.- 9.2 Research Design Components.- 9.3 Analysis of Results.- 9.4 Bias Assessment and Control.- 9.5 “Pseudo-Cohort” Cross-Sectional Studies.- 9.6 Advantages, Disadvantages, and Uses of Cross-Sectional Studies.- II Biostatistics.- 10: Introduction to Statistics.- 10.1 Variables.- 10.2 Populations, Samples, and Sampling Variation.- 10.3 Description vs Statistical Inference.- 10.4 Statistical vs Analytic Inference.- 11: Descriptive Statistics and Data Display.- 11.1 Continuous Variables.- 11.2 Categorical Variables.- 11.3 Concluding Remarks.- 12: Hypothesis Testing and P Values.- 12.1 Formulating and Testing a Research Hypothesis.- 12.2 The Testing of Ho.- 12.3 Type II Error and Statistical Power.- 12.4 Bayesian vs Frequentist Inference.- 13: Statistical Inference for Continuous Variables.- 13.1 Repetitive Sampling and the Central Limit Theorem.- 13.2 Statistical Inferences Using the t-Distribution.- 13.3 Calculation of Sample Sizes.- 13.4 Nonparametric Tests of Two Means.- 13.5 Comparing Three or More Means: Analysis of Variance.- 13.6 Control for Confounding Factors.- 14: Statistical Inference for Categorical Variables.- 14.1 Introduction to Categorical Data Analysis.- 14.2 Comparing Two Proportions.- 14.3 Statistical Inferences for a Single Proportion.- 14.4 Comparison of Three or More Proportions.- 14.5 Analysis of Larger (r × c) Contingency Tables.- 15: Linear Correlation and Regression.- 15.1 Linear Correlation.- 15.2 Linear Regression.- 15.3 Correlation vs Regression.- 15.4 Statistical Inference.- 15.5 Control for Confounding Factors.- 15.6 Rank (Nonparametric) Correlation.- III Special Topics.- 16: Diagnostic Tests.- 16.1 Introduction.- 16.2 Defining “Normal” and “Abnormal” Test Results.- 16.3 The Reproducibility and Validity of Diagnostic Tests.- 16.4 The Predictive Value of Diagnostic Tests.- 16.5 Bayes’ Theorem.- 16.6 The Uses of Diagnostic Tests.- 17: Decision Analysis.- 17.1 Strategies for Decision-Making.- 17.2 Constructing a Decision Tree.- 17.3 Probabilities and Utilities.- 17.4 Completing the Analysis.- 17.5 Cost-Benefit Analysis.- 17.6 Cost-Effectiveness Analysis.- 18: Life-Table (Survival) Analysis.- 18.1 Introduction.- 18.2 Alternative Methods of Analysis: an Example.- 18.3 The Actuarial Method.- 18.4 The Kaplan-Meier (Product-Limit) Method.- 18.5 Statistical Inference.- 19: Causality.- 19.1 What is a “Cause”?.- 19.2 Necessary, Sufficient, and Multiple Causes.- 19.3 Patterns of Cause.- 19.4 Probability and Uncertainty.- 19.5 Can Exposure Cause Outcome?.- 19.6 Is Exposure an Important Cause of Outcome?.- 19.7 Did Exposure Cause Outcome in a Specific Case?.- Appendix Tables.