"A Chapman & Hall book."

Includes bibliographical references (page 255-273) and index.

Machine generated contents note: 1.Introduction -- 1.1.Randomized Clinical Trials -- 1.2.Observational Studies -- 1.3.The Problem of Multiple Comparisons -- 1.4.The Evolution of Available Data Streams -- 1.5.The Hierarchy of Scientific Evidence -- 1.6.Statistical Significance -- 1.7.Summary -- 2.Basic Statistical Concepts -- 2.1.Introduction -- 2.2.Relative Risk -- 2.3.Odds Ratio -- 2.4.Statistical Power -- 2.5.Maximum Likelihood Estimation -- 2.5.1.Example with a Closed Form Solution -- 2.5.2.Example without a Closed Form Solution -- 2.5.3.Bayesian Statistics -- 2.5.4.Example -- 2.6.Non-linear Regression Models -- 2.7.Causal Inference -- 2.7.1.Counterfactuals -- 2.7.2.Average Treatment Effect -- 3.Multi-level Models -- 3.1.Introduction -- 3.2.Issues Inherent in Longitudinal Data -- 3.2.1.Heterogeneity -- 3.2.2.Missing Data -- 3.2.3.Irregularly Spaced Measurement Occasions -- 3.3.Historical Background --

Note continued: 3.4.Statistical Models for the Analysis of Longitudinal and/or Clustered Data -- 3.4.1.Mixed-effects Regression Models -- 3.4.1.1.Random Intercept Model -- 3.4.1.2.Random Intercept and Trend Model -- 3.4.2.Matrix Formulation -- 3.4.3.Generalized Estimating Equation Models -- 3.4.4.Models for Categorical Outcomes -- 4.Causal Inference -- 4.1.Introduction -- 4.2.Propensity Score Matching -- 4.2.1.Illustration -- 4.2.2.Discussion -- 4.3.Marginal Structural Models -- 4.3.1.Illustration -- 4.3.2.Discussion -- 4.4.Instrumental Variables -- 4.4.1.Illustration -- 4.5.Differential Effects -- 5.Analysis of Spontaneous Reports -- 5.1.Introduction -- 5.2.Proportional Reporting Ratio -- 5.2.1.Discussion -- 5.3.Bayesian Confidence Propagation Neural Network (BCPNN) -- 5.4.Empirical Bayes Screening -- 5.5.Multi-item Gamma Poisson Shrinker -- 5.6.Bayesian Lasso Logistic Regression -- 5.7.Random-effect Poisson Regression -- 5.7.1.Rate Multiplier -- 5.8.Discussion --

Note continued: 6.Meta-analysis -- 6.1.Introduction -- 6.2.Fixed-effect Meta-analysis -- 6.2.1.Correlation Coefficient -- 6.2.2.Mean Difference -- 6.2.3.Relative Risk -- 6.2.3.1.Inverse Variance Method -- 6.2.3.2.Mantel-Haenszel Method -- 6.2.4.Odds Ratio -- 6.2.4.1.Inverse Variance Method -- 6.2.4.2.Mantel-Haenszel Method -- 6.2.4.3.Peto Method -- 6.3.Random-effect Meta-analysis -- 6.3.1.Sidik-Jonkman Estimator of Heterogeneity -- 6.3.2.DerSimonian-Kacker Estimator of Heterogeneity -- 6.3.3.REML Estimator of Heterogeneity -- 6.3.4.Improved PM Estimator of Heterogeneity -- 6.3.5.Example -- 6.3.6.Issues with the Weighted Average in Meta-analysis -- 6.4.Maximum Marginal Likelihood/Empirical Bayes Method -- 6.4.1.Example: Percutaneous Coronary Intervention Based Strategy versus Medical Treatment Strategy -- 6.5.Bayesian Meta-analysis -- 6.5.1.WinBugs Example -- 6.6.Confidence Distribution Framework for Meta-analysis -- 6.6.1.The Framework -- 6.6.1.1.Fixed-effects Model --

Note continued: 6.6.1.2.Random-effects Model -- 6.6.2.Meta-analysis of Rare Events under the CD Framework -- 6.7.Discussion -- 7.Ecological Methods -- 7.1.Introduction -- 7.2.Time Series Methods -- 7.2.1.Generalized Event Count Model -- 7.2.2.Tests of Serial Correlation -- 7.2.3.Parameter-driven Generalized Linear Model -- 7.2.4.Autoregressive Model -- 7.3.State Space Model -- 7.4.Change-point Analysis -- 7.4.1.The u-chart -- 7.4.2.Estimation of a Change-point -- 7.4.3.Change-point Estimator for the INAR(1) Model -- 7.4.3.1.Change-point Estimator for the Rate Parameter -- 7.4.3.2.Change-point Estimator for the Dependence Parameter -- 7.4.4.Change-point of a Poisson Rate Parameter with Linear Trend Disturbance -- 7.4.5.Change-point of a Poisson Rate Parameter with Level and Linear Trend Disturbance -- 7.4.6.Discussion -- 7.5.Mixed-effects Poisson Regression Model -- 8.Discrete-time Survival Models -- 8.1.Introduction -- 8.2.Discrete-time Ordinal Regression Model --

Note continued: 8.3.Discrete-time Ordinal Regression Frailty Model -- 8.4.Illustration -- 8.5.Competing Risk Models -- 8.5.1.Multinomial Regression Model -- 8.5.2.Mixed-Effects Multinomial Regression Model -- 8.6.Illustration -- 8.6.1.Model Parameterization -- 8.6.2.Results -- 8.6.3.Discussion -- 9.Research Synthesis -- 9.1.Introduction -- 9.2.Three-level Mixed-effects Regression Models -- 9.2.1.Three-level Linear Mixed Model -- 9.2.1.1.Illustration: Efficacy of Antidepressants -- 9.2.2.Three-level Non-linear Mixed Model -- 9.2.3.Three-level Logistic Regression Model for Dichotomous Outcomes -- 9.2.3.1.Illustration: Safety of Antidepressants -- 10.Analysis of Medical Claims Data -- 10.1.Introduction -- 10.2.Administrative Claims -- 10.3.Observational Data -- 10.4.Experimental Strategies -- 10.4.1.Case-control Studies -- 10.4.2.Cohort Studies -- 10.4.3.Within-subject Designs -- 10.4.3.1.Self-controlled Case Series -- 10.4.4.Between-subject Designs --

Note continued: 10.5.Statistical Strategies -- 10.5.1.Fixed-effects Logistic and Poisson Regression -- 10.5.2.Mixed-effects Logistic and Poisson Regression -- 10.5.3.Sequential Testing -- 10.5.4.Discrete-time Survival Models -- 10.5.5.Stratified Cox Model -- 10.5.6.Between and Within Models -- 10.5.7.Fixed-effect versus Random-effect Models -- 10.6.Illustrations -- 10.6.1.Antiepileptic Drugs and Suicide -- 10.6.2.Description of the Data, Cohort, and Key Design and Outcome Variables -- 10.6.3.Statistical Methods -- 10.6.4.Between-subject Analyses -- 10.6.5.Within-subject Analysis -- 10.6.6.Discrete-time Analysis -- 10.6.7.Propensity Score Matching -- 10.6.8.Self-controlled Case Series and Poisson Hybrid Models -- 10.6.9.Marginal Structural Models -- 10.6.10.Stratified Cox and Random-effect Survival Models -- 10.6.11.Conclusion -- 10.7.Conclusion -- 11.Methods to be Avoided -- 11.1.Introduction -- 11.2.Spontaneous Reports -- 11.3.Vote Counting --

Note continued: 11.4.Simple Pooling of Studies -- 11.5.Including Randomized and Non-randomized Trials in Meta-analysis -- 11.6.Multiple Comparisons and Biased Reporting of Results -- 11.7.Immortality Time Bias -- 12.Summary and Conclusions -- 12.1.Final Thoughts.