Analysis of Incidence Rates

Incidence rates are counts divided by person-time, while mortality rates are a well-known example. Analysis of Incidence Rates discusses the practical aspects of analyzing incidence rates, including pitfalls and areas of controversy. It is aimed at graduate students, researchers, and analysts in epidemiology, biostatistics, social sciences, economics, and psychology. The text covers methods such as standardization, inverse-variance weighting, Poisson regression, linear regression, confidence intervals, collapsibility, extensions of count models, matched cohort methods, marginal methods, instrumental variable methods, and Royston-Parmar models.

Format: Hardback
Length: 474 pages
Publication date: 11 April 2019
Publisher: Taylor & Francis Ltd

Incidence rates are counts divided by person-time, while mortality rates are a well-known example. Analysis of Incidence Rates provides a comprehensive discussion of the practical aspects of analyzing incidence rates. It addresses important pitfalls and areas of controversy, making it valuable for graduate students, researchers, and analysts in epidemiology, biostatistics, social sciences, economics, and psychology.

The text covers various topics, including:

Comparison and Contrast of Incidence Rates with Risks, Odds, and Hazards: The text explains the differences between incidence rates, risks, odds, and hazards. It highlights the importance of understanding these concepts in epidemiological analysis.

Stratified Methods: The book discusses several stratified methods for analyzing incidence rates, such as standardization, inverse-variance weighting, and Mantel-Haenszel methods. It explains how these methods can be used to adjust for confounding factors and estimate rate differences between groups.

Poisson Regression Methods: Poisson regression is a commonly used method for analyzing incidence rates. The text provides an in-depth explanation of Poisson regression methods, including adjusted rate ratios and rate differences. It discusses the use of Poisson regression for both binary and continuous outcomes.

Linear Regression for Rate Differences: Linear regression is used to estimate rate differences between groups when the outcome is continuous. The text discusses linear regression methods, including common problems such as collinearity and heteroscedasticity. It also provides methods for correcting confidence intervals and exploring extensions of count models for rates.

Correcting Confidence Intervals: Confidence intervals are used to estimate the uncertainty of incidence rate estimates. The text discusses methods for correcting confidence intervals, including the use of bootstrapping and permutation tests. It also explores problems related to collapsibility and presents methods for addressing them.

Extensions of Count Models for Rates: The book discusses extensions of count models for rates, including negative binomial regression, methods for clustered data, and the analysis of longitudinal data. It also reviews controversies and limitations associated with these models.

Matched Cohort Methods: Matched cohort methods are used to estimate the effect of an exposure on incidence rates. The text provides a detailed explanation of matched cohort methods, including the selection of cohorts, matching criteria, and estimation of effect sizes.

Marginal Methods for Converting Adjusted Rate Ratios to Rate Differences and Vice Versa: Marginal methods are used to convert adjusted rate ratios to rate differences and vice versa. The text discusses these methods, including the use of Stata commands and examples.

Instrumental Variable Methods: Instrumental variable methods are used to address the problem of confounding when the exposure and outcome are associated with each other. The text discusses instrumental variable methods, including the use of instrumental variables and their limitations.

Comparison of Poisson Regression with the Cox Proportional Hazards Model: The Cox proportional hazards model is another commonly used method for analyzing incidence rates. The text compares Poisson regression with the Cox proportional hazards model, discussing their advantages and limitations.

Royston-Parmar Models: Royston-Parmar models are used to estimate incidence rates in the presence of competing risks. The text introduces Royston-Parmar models and discusses their applications in epidemiological analysis.

All data and analyses are provided in online Stata files, allowing readers to download and explore the data themselves.

Peter Cummings, Professor Emeritus at the Department of Epidemiology, School of Public Health, University of Washington, Seattle, WA, is the author of Analysis of Incidence Rates. His research primarily focused on injuries, and he used matched cohort methods to estimate the impact of seat belts and the presence of airbags on injury outcomes.

In conclusion, Analysis of Incidence Rates is a comprehensive and practical guide to analyzing incidence rates in epidemiological research. It covers a wide range of topics, including comparison and contrast of incidence rates with risks, odds, and hazards, stratified methods, Poisson regression methods, linear regression for rate differences, correcting confidence intervals, extensions of count models for rates, matched cohort methods, marginal methods, instrumental variable methods, comparison of Poisson regression with the Cox proportional hazards model, and Royston-Parmar models. The book is aimed at graduate students, researchers, and analysts in the disciplines of epidemiology, biostatistics, social sciences, economics, and psychology, and it provides valuable insights and techniques for conducting effective incidence rate analysis.

Weight: 1124g
Dimension: 184 x 260 x 36 (mm)
ISBN-13: 9780367152062