Misclassification of Smoking Habits and Passive Smoking

How accurate are statements about smoking habits? This book presents the results of a comprehensive review in which the literature on the subject is newly interpreted. It is shown that smokers are misclassified as non-smokers in epidemiological studies often enough to explain the increased lung cancer risk seen in self-reported non-smokers in relation to their spouse's smoking habits. This study overturns the commonly held view that increased risk is a consequence of exposure to environmental tobacco smoke and highlights the difficulty of making valid inferences from epidemiological data. No-one should draw conclusions about passive smoking before reading this book!

Inhaltsangabe1 Introduction.- 1.1 Objectives.- 1.2 Reasons for Study of Inaccuracy of Smoking Habits.- 1.3 Epidemiological Evidence of an Association of Passive Smoking and Lung Cancer.- 1.4 Possible Explanations for the Association.- 1.4.1 Non-Reporting Bias.- 1.4.2 Passive Smoke Exposure Recall Bias.- 1.4.3 Misclassification of Active Smoking Habits.- 1.5 First Attempts to Estimate Misclassification Bias.- 1.6 Difference in Conclusions and Need for More Data.- 1.7 Misclassification Not the Only Issue.- 1.8 Structure of the Monograph.- 2 General Considerations.- 2.1 Major and Minor Sources of Misclassification Bias in Passive Smoking Studies.- 2.2 Types of Information on Misclassification of Smoking Habit.- 2.2.1 Information from 2 Sources at the Individual Level.- 2.2.2 One Source of Information Assumed Correct.- 2.2.3 Neither Source of Information Necessarily Correct.- 2.2.4 Types of Situation Where Two Sources of Information are Available.- 2.2.5 Information from 2 Sources at the Group Level.- 2.3 Factors Which Might Affect Reported Smoking Habits.- 2.4 Use of Objective Markers.- 2.4.1 General Problems.- 2.4.2 Carbon Monoxide and Carboxyhaemoglobin.- 2.4.3 Thiocyanate.- 2.4.4 Nicotine and Cotinine.- 3 Evidence on Misclassification.- 3.1 Studies Using Carbon Monoxide and Thioeyanate as Objective Markers.- 3.1.1 Overview.- 3.1.2 The Studies.- 3.2 Studies Using Nicotine or Cotinine as Objective Markers.- 3.2.1 Overview.- 3.2.2 The Studies.- 3.3 Studies Involving Multiple Reports Concerning the Same Individual - Subject a Lung Cancer Case.- 3.3.1 Overview.- 3.3.2 The Studies.- 3.4 Studies Involving Multiple Reports Concerning the Same Individual - Subject not a Lung Cancer Case.- 3.4.1 Overview.- 3.4.2 The Studies.- 3.5 Other Evidence on Misclassification of Smoking Habits.- 3.5.1 Overview.- 3.5.2 The Studies.- 4 Overview and Unanswered Questions.- References.- Appendix A Studies of Passive Smoking and Lung Cancer - A Brief Review with Special Reference to the Way the Smoking Habit was Obtained.- Appendix B Passive Smoking and Lung Cancer Association: A Result of Bias?.- Appendix C Major and Minor Sources of Misclassification Bias in Passive Smoking Studies - Mathematical Details.- Appendix D Trends in the Percentages of UK Men and Women who Have Never Smoked.