Epidemiology and Biostatistics: An Introduction to Clinical Research

Description

This is a concise introduction to epidemiology and biostatistics written specifically for medical students and first-time learners of clinical research methods.  It presents the core concepts of epidemiology and of biostatistics and illustrates them with extensive examples from the clinical literature.  It is the only book on the market written to speak directly to medical students and first-time biomedical researchers by using language and examples that are easy to understand.  This newly updated second edition is extensively rewritten to provide the clearest explanations and examples. There is also a sister-text, a 150-problem workbook of practice problems that can be purchased alongside this textbook.  The author continues to provide a text that is attractively fast-paced and concise for use in condensed courses, such as those taught in medical school.  The book is an excellent review for the epidemiology section of the United States Medical Licensing Examination Part I which all medical students must take at the end of the second year. Preface Contents Part I: Epidemiology Chapter 1: Causal Relationships in€Health and€Disease 1.1 Inferring Causation from€Epidemiologic Studies 1.2 Factors Favoring an€Inference of€Causation 1.2.1 Evidence Arising from€Randomized Studies 1.2.2 Strength of€Association 1.2.3 Temporal Relationship 1.2.4 Exposure-Varying Association 1.2.5 Biological Plausibility References Chapter 2: Basic Measures of€Disease Frequency 2.1 Prevalence 2.1.1 Definition of€Prevalence 2.1.2 Applications of€Prevalence Data 2.1.3 Limitation of€Prevalence Measures 2.2 Incidence 2.2.1 Definitions of€Incidence 2.2.2 Applications of€Incidence Data 2.3 Relationship Between Prevalence and€Incidence 2.4 Stratification of€Disease Frequencies by Person, Place, and€Time 2.4.1 Measures of€Disease Frequency Stratified by€Characteristics of€Person 2.4.2 Measures of€Disease Frequency Stratified by€Characteristics of€Place 2.4.3 Measures of€Disease Frequency Stratified by€Characteristics of€Time References Chapter 3: General Considerations in€Epidemiologic€Research 3.1 Interventional Versus Observational Study Designs 3.1.1 Interventional Studies Can Isolate the€Causal Impact of€Specific Treatments 3.1.2 Interventional Studies Are Limited to€Evaluation of€Specific Treatments and€Diseases 3.1.3 The Results of€Interventional Studies May Have Limited Applicability 3.2 Study Population 3.2.1 Source Population 3.2.2 Exclusion Criteria 3.2.2.1 Exclusion of€People Who Have Prevalent Disease 3.2.2.2 Exclusion of€People Who Have Another Strong Risk Factor for€the€Disease 3.2.2.3 Exclusion of€People Who Cannot Provide Reliable Study Data 3.2.2.4 Exclusion of€People Who Cannot Complete the€Study Safely 3.2.3 Where to€Find Information About the€Study Population in€a€Research Article 3.3 Exposure and€Outcome 3.3.1 Definition 3.3.2 Measuring the€Study Data 3.3.3 Where to€Find Information About the€Exposure and€Outcome in€a€Research Article 3.4 Internal and€External Validity 3.5 Summary of€Common Research Study Designs Reference Chapter 4: Case Reports and€Case Series References Chapter 5: Cross-Sectional Studies Reference Chapter 6: Cohort Studies 6.1 Cohort Study Design 6.1.1 Exclusion for€Prevalent Disease 6.1.2 Creation of€the€Cohorts 6.1.3 Determination of€the€Outcome 6.2 Quality of€the€Exposure Measurements 6.2.1 Are the€Measurements Accurate? 6.2.2 Are the€Measurements Precise? 6.2.3 Are the€Measurements Applied Impartially to€the€Study Population? 6.2.4 Are the€Measurements Performed at the€Right Time? 6.2.5 Retrospective Versus Prospective Data Collection 6.3 Pharmacoepidemiology Studies 6.4 Analysis of€Cohort Study Data 6.4.1 Calculation of€Disease Incidences Among the€Cohorts 6.4.2 Comparison of€Disease Incidences Among the€Cohorts 6.4.2.1 Relative Risk 6.4.2.2 Attributable Risk 6.4.2.3 Population Attributable Risk 6.4.2.4 Other Measures of€Risk 6.5 Advantages of€Cohort Studies 6.5.1 Ability to€Discern Temporal Relationships Between Exposure and€Disease 6.5.2 Ability to€Study Multiple Outcomes 6.6 Limitations of€Cohort Studies 6.6.1 Confounding 6.6.2 Inefficient Design for€Rare Diseases and€Those with€a€Long Latency Period References Chapter 7: Case-Control Studies 7.1 Case-Control Study Design 7.2 Selection of€Cases and€Controls 7.2.1 Select Case Individuals Using a€Specific Definition of€the€Disease 7.2.2 Select Case Individuals Close to€the€Time of€Initial Disease Development 7.2.3 Select Control Individuals from€the€Same Underlying Population as€the€Cases 7.2.4 Select Control Individuals Who Have the€Same Opportunity to€Be€Counted as€a€Case 7.2.5 Nested Case-Control Studies 7.2.6 Matching 7.2.7 Number of€Controls 7.3 Analysis of€Case Control Study Data 7.3.1 Concept of€the€Odds Ratio 7.3.2 Practical Calculation of€the€Odds Ratio 7.3.3 Odds Ratios and€Relative Risk 7.4 Advantages of€Case Control Studies 7.4.1 Ideal for€Studying Rare Diseases and€Those with€a€Long Latency Period 7.4.2 Efficiency 7.4.3 Evaluation of€Multiple Exposures 7.5 Limitations of€Case Control Studies 7.5.1 Confounding 7.5.2 Requires Ascertainment of€Previous Exposures in€Retrospect 7.5.3 Inability to€Directly Determine the€Incidence of€Disease References Chapter 8: Randomized Trials 8.1 Rationale for€Randomized Trials 8.2 General Design of€Randomized Trials 8.3 Trial Populations 8.3.1 Definition of€the€Target Condition 8.3.2 Exclusion of€People Suspected to€Have Difficulty Adhering to€the€Study Treatments 8.3.3 Exclusion of€People Who Have Comorbid Conditions 8.3.4 Exclusion of€People Who Are Already Receiving the€Study Treatment 8.3.5 Exclusion for€Safety 8.3.6 Broadly Inclusive Healthcare Settings Promote Applicability of€Trial Results 8.4 Interventions and€Control Procedures 8.4.1 Intervention 8.4.2 Control Procedures 8.4.2.1 Placebo 8.4.2.2 Delayed Treatment 8.4.2.3 Pre-existing Standard of€Care 8.4.2.4 No Treatment 8.5 Outcomes of€Trials 8.5.1 Capture Important Benefits and€Harms of€the€Intervention 8.5.2 Types of€Outcomes 8.5.3 Mortality Outcomes 8.5.4 Surrogate Outcomes 8.5.5 Premature Termination of€Trials 8.6 Procedures to€Promote Internal Validity of€Trials 8.6.1 Randomization 8.6.2 Blinding 8.6.3 Concealment of€Treatment Allocation 8.6.4 Efficacy and€Effectiveness 8.6.5 Trial Reporting 8.7 Specific Trial Designs 8.7.1 Factorial Trials 8.7.2 Crossover Trials 8.7.3 Phases of€Drug Development 8.7.3.1 Phase I€Studies 8.7.3.2 Phase II Studies 8.7.3.3 Phase III Studies 8.7.3.4 Phase IV Studies 8.8 Analysis of€Clinical Trial Data 8.8.1 Measures of€Effect 8.8.1.1 Relative Risk 8.8.1.2 Attributable Risk 8.8.1.3 Number Needed to€Treat or Harm 8.8.1.4 Changes in€Quantitative Outcomes 8.8.2 Intention-To-Treat Analysis 8.8.3 Subgroup Analyses 8.9 Limitations of€Randomized Trials 8.9.1 Limited External Validity (Applicability) of€the€Trial Population 8.9.2 Limited External Validity (Applicability) of€the€Trial Environment 8.9.3 Narrow Study Question 8.9.4 Randomized Design Accounts Only for€Confounding 8.9.5 Negative Trials References Chapter 9: Misclassification 9.1 Definition of€Miscalculation 9.2 Non-differential Miscalculation 9.2.1 Non-differential Misclassification of€the€Exposure 9.2.2 Non-differential Misclassification of€the€Outcome 9.2.3 Summary of€Non-differential Misclassification 9.3 Differential Misclassification 9.3.1 Differential Misclassification of€the€Exposure 9.3.2 Differential Misclassification of€the€Outcome 9.3.3 Summary of€Differential Misclassification 9.4 Assessment of€Misclassification in€Research Articles Reference Chapter 10: Confounding 10.1 Confounding Obscure Understanding of€Causal Relationships 10.2 Evaluation of€a€Potential Confounding Characteristic 10.2.1 Confounding Characteristic Associated with€the€Exposure 10.2.2 Confounding Characteristic Associated with€the€Outcome 10.2.3 Confounding Characteristic Not on€the€Causal Pathway of€Association 10.2.4 Evaluation of€Confounding Characteristics in€Research Articles 10.3 Residual Confounding 10.4 Confounding by Indication 10.5 Methods to€Control for€Confounding 10.5.1 Method of€Restriction 10.5.2 Method of€Stratification plus Adjustment 10.5.3 Method of€Matching 10.5.4 Method of€Regression 10.6 Interpreting Results After Adjustment for€Confounding Reference Chapter 11: Effect Modification 11.1 Concept of€Effect Modification 11.2 Evaluation of€Effect Modification 11.2.1 Criteria for€Assessing the€Presence of€Effect Modification 11.2.2 Statistical Considerations for€Evaluating Effect Modification 11.2.2.1 Confidence Intervals 11.2.2.2 P-value for€Interaction 11.2.2.3 Multiple Testing 11.3 Effect Modification and€Confounding Are Distinct Concepts 11.4 Effect Modification on€the€Relative and€Absolute Scales References Chapter 12: Screening and€Diagnosis 12.1 General Principles of€Screening and€Diagnosis 12.2 Utility of€Testing 12.3 Qualities of€Diseases Appropriate for€Screening 12.3.1 Early Recognition of€the€Disease Should Provide Meaningful Benefit 12.3.2 Screening Tests Should Target Diseases that Have Potentially Serious Consequences 12.3.3 Diseases Targeted by Screening Require a€Preclinical Phase 12.4 Qualities of€Tests Appropriate for€Screening or Diagnosis 12.4.1 Validity 12.4.1.1 Sensitivity and€Specificity 12.4.1.2 Positive and€Negative Predictive Values 12.4.1.3 Likelihood Ratios 12.4.2 Reliability 12.4.2.1 Coefficient of€Variation 12.4.2.2 Kappa 12.5 Defining Cut Points for€Continuous Tests 12.6 Types of€Biases in€Screening Studies 12.6.1 Confounding (Referral Bias) 12.6.2 Lead Time Bias 12.6.3 Length Bias Sampling 12.7 Levels of€Prevention 12.7.1 Primary Prevention 12.7.2 Secondary Prevention 12.7.3 Tertiary Prevention 12.8 Association Is Not Sufficient for€Prediction References Part II: Biostatistics Chapter 13: Summary Measures in€Statistics 13.1 Types of€Variables 13.2 Univariate Statistics 13.2.1 Histograms 13.2.2 Measures of€Location and€Spread 13.2.3 Quantiles 13.2.4 Univariate Statistics for€Binary Data 13.3 Bivariate Statistics 13.3.1 Tabulation Across Categories 13.3.2 Correlation 13.3.3 Quantile-Continuous Variable Plots Chapter 14: Introduction to€Statistical Inference 14.1 Definition of€a€Population and€a€Sample 14.2 External Validity 14.3 Statistical Inference 14.4 Confidence Intervals 14.5 Hypothesis Testing 14.5.1 Construction of€Statistical Hypotheses 14.5.2 P-Values Reference Chapter 15: Hypothesis Tests in€Practice 15.1 Two-Sample Hypothesis Tests 15.1.1 T-Test 15.1.2 Chi-Square Test 15.1.3 Hypothesis Tests for€Other Types of€Study Data 15.1.4 Multiple Sample Hypothesis Tests 15.2 An Imperfect System 15.2.1 Type I€Error 15.2.2 Type II Error 15.3 Power 15.3.1 Sample Size (N) 15.3.2 Effect Size (μ) 15.3.3 Variability (σ) 15.3.4 Significance Level (α) Chapter 16: Linear Regression 16.1 Describing the€Association Between Two Variables 16.2 Univariate Linear Regression 16.2.1 The Linear Regression Equation 16.2.2 Residuals and€the€Sum of€Squares 16.2.3 Interpreting Continuous Covariates from€a€Linear Regression Model 16.2.4 Interpreting Binary Covariates from€Linear Regression Equations 16.3 Diagnostics 16.3.1 Absolute Versus Relative Fit 16.3.2 Nonlinear Associations 16.3.3 Influential Points 16.3.4 Extrapolating the€Regression Equation Beyond the€Observed Data 16.4 Multiple Linear Regression 16.4.1 Definition of€the€Multiple Regression Model 16.4.2 Interpreting Results from€the€Multiple Regression Model 16.4.2.1 Obtaining Estimated Values for€a€Particular Set of€Data 16.4.2.2 Obtaining Relative Differences in€the€Outcome Variable by a€Covariate 16.4.2.3 Interpreting Multiple Regression Results in€Research Articles 16.4.2.4 P-Values and€Confidence Intervals for€Regression Coefficients 16.5 Confounding and€Effect Modification in€Multiple Regression Models 16.5.1 Confounding 16.5.2 Effect Modification Chapter 17: Log-Link and€Logistic Regression 17.1 Regression for€Ratios 17.1.1 Log-Link Regression 17.1.2 Interpretation of€Log-Link Regression Models 17.1.3 Hypothesis Testing for€Log-Link Regression Results 17.2 Logistic Regression 17.2.1 Definition and€Interpretation of€the€Logistic Regression Model 17.2.2 Interpreting Logistic Regression Results from€Research Articles Chapter 18: Survival Analysis 18.1 Motivation for€Survival Data 18.2 Interpretation of€Survival Data 18.2.1 Description of€the€Survivor Function 18.2.2 Estimating Time-Specific and€Median Survival from€S(t) 18.2.3 Statistical Testing of€Survival Data 18.3 Estimation of€the€Survivor Function 18.3.1 Definitions of€Outcomes and€Censoring 18.3.2 Kaplan-Meier Estimation of€the€Survivor Function for€Uncensored Data 18.3.3 Kaplan-Meier Estimation of€the€Survivor Function for€Censored Data 18.4 Cox’s Proportional Hazards Model 18.5 Interpreting Survival Data 18.5.1 Interpreting Hazard Ratios 18.5.2 Interpreting Survival Versus Hazard Ratio Data Reference Glossary of€Terms Index