Statistics for Dental Clinicians

Preamble xi 1 What is statistics and why do we need it? 1 Selected readings 7 2 Understanding and interpreting measures of association 8 Effect and effect size 8 Dichotomous outcome variables 9 Data presentation and interpretation 9 Absolute risk 9 Absolute risk difference 10 Relative risk or risk ratio 10 Odds ratio 11 Mean difference 11 A few additional notes 12 Selected readings 16 3 Understanding and interpreting a standard deviation and normal distribution 17 A few additional notes 22 Selected readings 23 4 Understanding and interpreting standard error 24 Interpretation of the standard error of the mean 25 Implication of the standard error of the mean 26 A few additional notes 26 Selected readings 29 5 Understanding and interpreting hypothesis testing and p-values 30 Descriptive and inferential statistics 30 Sampling error 31 Hypothesis testing or null hypothesis significance testing 31 Null and alternative hypotheses 32 Significance 33 P-value 34 A few additional notes 36 Selected readings 37 6 Understanding and interpreting a confidence interval 38 Understanding the confidence interval 39 How to interpret a confidence interval 39 A few additional notes 41 Selected readings 44 7 Understanding and interpreting power analysis and sample size 45 Sample size: Why is it important? 45 Components of a sample size calculation 46 Size of the effect 46 Significance level and type I error 47 Power and type II error 49 A few additional notes 50 Selected readings 53 8 Understanding and interpreting a survival analysis 54 Kaplan-Meier or survival curve 55 Comparing two Kaplan-Meier (survival) curves 57 Cox proportional hazard model 58 A few additional notes 59 Selected readings 60 9 Understanding and interpreting a probabilistic-based diagnosis 61 Sensitivity 61 Specificity 63 Positive predictive value 63 Negative predictive value 63 Likelihood ratios 65 Selective readings 68 10 Understanding and interpreting a correlation 69 Pearson product-moment correlation 69 Interpretation of Pearson correlation coefficients and coefficient of determination 71 Misinterpretation of correlations 73 A few additional notes 74 Selected readings 74 11 Understanding and interpreting a regression analysis 76 Estimation 76 Prediction 77 Linear regression 77 Multiple (or multivariable) linear (MLR) regression 79 Logistic regression 79 Predicting risk and odds 80 A hypothetical example: predicting risk and odds of an outcome 80 Estimating odds ratios 81 A hypothetical example--estimating an odds ratio 81 Nonindependence of observations 82 Building a regression model 82 A few additional notes 82 Selected readings 85 12 Understanding and interpreting confounding and effect modification 86 Counterfactual framework and causal reasoning 86 Causal inference and confounding bias 87 Strategies to deal with confounding in the study design phase 89 1. Randomization 90 2. Specification 90 3. Matching 90 Strategies to deal with confounding in the analysis phase 92 1. Stratification 92 2. Propensity score 92 3. Traditional regression modeling 93 Effect modification 94 A few additional notes 95 Selected readings 97 13 Understanding and interpreting bias 98 Random error versus systematic error (bias) 98 What does it really mean? 99 Distinguishing risk of bias, methodological quality, and reporting quality 100 Assessing risk of bias in primary studies 101 A few additional notes 104 Selected readings 105 14 Understanding and interpreting patient-reported outcomes 106 Identifying optimal patient-reported outcome measures 108 Validity 108 Reliability 109 A hypothetical scenario--Cohen's kappa (k) 109 Responsiveness 110 A few additional notes 111 Selected readings 112 15 Understanding and interpreting a cross-sectional study 113 Bias in cross-sectional studies 113 Response rate and avoiding nonresponse 114 Analysis of cross-sectional studies 115 Advantages and limitations of a cross-sectional study 116 A few additional notes 116 Selected readings 119 16 Understanding and interpreting a case-control study 120 Selection of the study population 120 Identifying cases 120 Identifying controls 122 Retrospective assessment of the exposure 122 Strengths and limitations 124 Selected readings 125 17 Understanding and interpreting a cohort study 126 Types of cohort study designs 126 Selection of the study population 127 Measuring exposures 128 Measuring outcome frequency 129 Measures of association 129 Bias in cohort studies 130 Selection bias 130 Nonparticipation and nonresponse 130 Loss to follow-up or attrition bias 130 Information bias: Dissimilar information between exposed and unexposed participants 130 Confounding 131 Strengths and limitations 131 Suggested readings 132 18 Understanding and interpreting a randomized controlled trial 133 Study arms 135 Type of outcomes 135 Methodological strategies in randomized controlled trials 137 Nonadherence to study protocol 137 Missing data 139 Subgroup analysis or effect modification 140 A few additional notes 142 Selected readings 144 19 Understanding and interpreting meta-analyses 145 The value of meta-analysis 145 Pairwise meta-analysis 146 Fixed effect meta-analysis 146 Random effects meta-analysis 147 Weight of each study in a meta-analysis 147 Forest plots 147 Heterogeneity 149 Network meta-analysis 151 Sensitivity analysis 152 Certainty of the evidence 152 A few additional notes 153 Selected readings 154 20 Understanding and interpreting statistical and clinical significance 155 A few additional notes 158 Selected readings 158 Appendix 1 Formulas and equations 160 Appendix 2 Z-table 183 Appendix 3 T-table 185 Glossary 189 Index 210