Epidemiology: An IntroductionIn the past thirty years epidemiology has matured from a fledgling scientific field into a vibrant discipline that brings together the biological and social sciences, and in doing so draws upon disciplines ranging from statistics and survey sampling to the philosophy of science. These areas of knowledge have converged into a modern theory of epidemiology that has been slow to penetrate into textbooks, particularly at the introductory level. Epidemiology: An Introduction closes the gap. It begins with a brief, lucid discussion of causal thinking and causal inference and then takes the reader through the elements of epidemiology, focusing on the measures of disease occurrence and causal effects. With these building blocks in place, the reader learns how to design, analyze and interpret problems that epidemiologists face, including confounding, the role of chance, and the exploration of interactions. All these topics are layered on the foundation of basic principles presented in simple language, with numerous examples and questions for further thought. |
From inside the book
Results 1-5 of 88
Page 3
... results that appear to confer an advantage on the smokers ? It is theoretically possible that all or many of the smokers quit soon after the survey and that many of the nonsmokers started smoking . While possible , this scenario is ...
... results that appear to confer an advantage on the smokers ? It is theoretically possible that all or many of the smokers quit soon after the survey and that many of the nonsmokers started smoking . While possible , this scenario is ...
Page 4
... result is a strikingly different age distribution for the female smokers and non- smokers of Whickham . Were this difference in the age distribution ig- nored , one might conclude erroneously that smoking was not related to a higher ...
... result is a strikingly different age distribution for the female smokers and non- smokers of Whickham . Were this difference in the age distribution ig- nored , one might conclude erroneously that smoking was not related to a higher ...
Page 9
... result we slip into a frame of thinking in which we identify the switch as a unique cause . The inadequacy of this assumption is emphasized when the bulb goes bad and needs to be replaced . The Causal Pie Model Causes of disease can be ...
... result we slip into a frame of thinking in which we identify the switch as a unique cause . The inadequacy of this assumption is emphasized when the bulb goes bad and needs to be replaced . The Causal Pie Model Causes of disease can be ...
Page 10
... result from it , we can prevent the disease by appropriate dietary intervention . Thus , we can say that the disease ... results in a broken hip . Other factors playing a causal role for the broken hip could include the type of shoe the ...
... result from it , we can prevent the disease by appropriate dietary intervention . Thus , we can say that the disease ... results in a broken hip . Other factors playing a causal role for the broken hip could include the type of shoe the ...
Page 11
... result , the concept of a strong or weak cause cannot be a universally accurate description of any cause . For example , suppose we say that smoking is a strong cause of lung cancer because it plays a causal role in a large proportion ...
... result , the concept of a strong or weak cause cannot be a universally accurate description of any cause . For example , suppose we say that smoking is a strong cause of lung cancer because it plays a causal role in a large proportion ...
Contents
1 | |
8 | |
3 Measuring Disease Occurrence and Causal Effects | 24 |
4 Types of Epidemiologic Study | 57 |
5 Biases in Study Design | 94 |
6 Random Error and the Role of Statistics | 113 |
7 Analyzing Simple Epidemiologic Data | 130 |
8 Controlling Confounding by Stratifying Data | 144 |
9 Measuring Interactions | 168 |
10 Using Regression Models in Epidemiologic Analysis | 181 |
11 Epidemiology in Clinical Settings | 198 |
Appendix | 218 |
Index | 221 |
Other editions - View all
Common terms and phrases
age categories asbestos attributable fraction average baseline bias biologic interaction birth order breast cancer calculate case-control data case-control study causal causal mechanism Chapter cigarette smoking clozapine cohort study compared component causes confidence interval confidence limits control confounding control series crude data curve data in Table dence rate denominator described distribution drug epidemic epidemiologic epidemiologic study evaluation example experiment exposed and unexposed exposed group Figure flutamide follow-up formula imbalance inci incidence proportion incidence rate ratio inference laryngeal cancer leukemia lung cancer measure misclassification mortality rate myocardial infarction nonsmokers null hypothesis obtain occur odds ratio outcome patients person person-time person-years placebo pooled estimate population at risk predict prevalence random error randomized trial rate difference relation result risk data risk difference risk factor risk of death risk ratio Rothman sampling screening source population specific standard strata stratified analysis Suppose syndrome tamoxifen tion tolbutamide treatment value function variable women