My Account Log in

2 options

Modeling in medical decision making : a Bayesian approach / Giovanni Parmigiani.

Table of contents Available online

View online
Holman Biotech Commons R723.5 .P37 2002
Loading location information...

Available This item is available for access.

Log in to request item
Format:
Book
Author/Creator:
Parmigiani, G. (Giovanni)
Series:
Statistics in practice
Language:
English
Subjects (All):
Medicine--Decision making--Statistical methods.
Medicine.
Bayesian statistical decision theory.
Decision Support Techniques.
Bayes Theorem.
Decision Making, Computer-Assisted.
Models, Statistical.
Medicine--Decision making.
Statistics.
Medical Subjects:
Decision Support Techniques.
Bayes Theorem.
Decision Making, Computer-Assisted.
Models, Statistical.
Physical Description:
xi, 266 pages : illustrations ; 24 cm.
Place of Publication:
Chichester ; New York, NY : J. Wiley, [2002]
Contents:
Part I Methods 1
1 Inference 3
1.2 Medical diagnosis 5
1.2.1 Sensitivity, specificity and positive predictive value 5
1.2.2 Bayes' rule 9
1.2.3 Odds, Bayes factors and the nomogram 11
1.3 Genetic counseling 15
1.3.1 Genetic susceptibility to breast and ovarian cancer 15
1.3.2 Individualized probabilities 16
1.3.3 Genetic parameters 18
1.3.4 Bayes' rule in genetic counseling 20
1.3.5 Conditional independence 22
1.3.6 Bayes factors for family history 23
1.3.7 Sequential use of Bayes' rule 29
1.4 Estimating sensitivity and specificity 31
1.4.1 Population and statistical model 31
1.4.2 Prior and posterior distributions 33
1.4.3 Distributions of positive and negative predictive values 37
1.4.4 Beta prior distributions 38
1.4.5 Prior-data conflict 40
1.4.6 Future patients 44
1.5 Chronic disease modeling 46
1.5.1 A simple Markov process for strokes - the [mu]SPPM 46
1.5.2 Estimating transition models from cohort data 49
2 Decision making 51
2.2 Foundations of expected utility theory 52
2.2.1 Formalizing health interventions and their consequences 53
2.2.2 The expected utility principle 54
2.2.3 Historical background 55
2.2.4 Axiomatic foundations 57
2.3 Measuring the value of avoiding a major stroke 59
2.3.1 Elicitation of personal utilities using the standard gamble approach 60
2.3.2 Quality-adjusted life years and the time trade-off method 62
2.3.3 Relationship between the standard gamble and the time trade-off method 64
2.3.4 Combining time trade-off and standard gamble to elicit utilities of risk-averse patients 65
2.3.5 Using individual utilities in medical decisions 67
2.4 Decision making in health care 68
2.4.1 Evidence-based medicine and synthesis of evidence 68
2.4.2 Elements of a decision model 71
2.4.3 Variation and uncertainty in decision modeling 73
2.4.4 Expected utilities 75
2.4.5 Plug-in approximation to expected utility 77
2.5 Cost-effectiveness analyses in the [mu]SPPM 78
2.5.1 From decision modeling to cost-effectiveness 78
2.5.2 The [mu]SPPM example 80
2.6 Statistical decision problems 82
2.6.1 Statistical decision theory 82
2.6.2 Loss functions 83
2.6.3 Implications of the expected utility principle for statistical reasoning 84
2.6.4 Hypothesis testing 85
2.6.5 Medical diagnosis 88
2.6.6 Choosing the number of subjects in a study 89
3 Simulation 91
3.2 Inference via simulation 92
3.2.1 Inference on sensitivity and specificity without a gold standard 92
3.2.2 Sampling from the posterior distribution 95
3.2.3 Markov chain Monte Carlo methods 99
3.3 Prediction and expected utility via simulation 103
3.3.1 Generating synthetic cohorts 103
3.3.2 Generating a synthetic cohort in the [mu]SPPM 104
3.4 Sensitivity analysis via simulation 105
3.4.1 Probabilistic sensitivity analysis 105
3.4.2 Probabilistic sensitivity analysis of the MCER in the [mu]SPPM 107
3.5 Searching for strategies via simulation 111
3.5.1 A two-phase design in screening for a rare disease 111
3.5.2 Monte Carlo evaluation of the expected utility 113
3.5.3 Markov chain sampling of actions 116
4 Meta-analysis 123
4.2 Meta-analysis 123
4.3 Bayesian meta-analysis 126
4.4 Tamoxifen in early breast cancer 127
4.4.2 Modeling heterogeneity 129
4.4.3 Computing 133
4.4.4 Results 136
4.5 Combining studies with continuous and dichotomous responses 138
4.5.1 A simulated example 139
4.6 Migraine headache 142
4.6.1 Background and goals 142
4.6.2 Data 144
4.6.3 Sampling distributions and latent variables 146
4.6.4 Treatment and study variation 149
4.6.5 Prior distributions 152
4.6.6 Computing 153
4.6.7 Results 154
4.6.8 Comparisons with alternative approaches 157
4.6.9 Graphical model checking 159
4.6.10 Sensitivity to prior specification 163
5 Decision trees 167
5.2 Axillary lymph node dissection in early breast cancer 167
5.2.1 Why a decision model? 169
5.3 A simple decision tree 170
5.3.1 Nodes, branches, and leaves 170
5.3.2 Backward induction 173
5.3.3 Measuring the value of prognostic information 174
5.3.4 The value of information 176
5.4 A more complete decision tree for ALND 177
5.4.1 Tree structure 177
5.4.2 Population model 178
5.4.3 Quality of life 180
5.4.4 Solving the tree for a specific patient type 182
5.4.5 Mapping the value of ALND in the covariates space 185
5.4.6 Uncertainty and probabilistic sensitivity analysis 189
6 Chronic disease modeling 195
6.2 Model overview 195
6.2.1 Modeling in breast cancer screening 195
6.2.2 Model components 196
6.3 Natural history model 198
6.3.1 States and transitions 198
6.3.2 Transition densities and sojourn time distribution 200
6.3.3 Estimating transition densities via deconvolution 202
6.3.4 Alternative estimates of preclinical sojourn time 204
6.3.5 Age of onset of preclinical disease 207
6.3.6 Other risk factors 210
6.4 Modeling the effects of screening 210
6.4.1 Sensitivity model 210
6.4.2 Axillary lymph node involvement model 211
6.4.3 Survival model 212
6.4.4 Costs 213
6.5 Comparing screening schedules 213
6.5.1 Screening strategies 213
6.5.2 Balance sheet 215
6.5.3 Trade-off analysis 216
6.6 Model critique 219
6.6.1 Limitations of the model 219
6.6.2 Clinical trials and screening women in their forties 221
6.6.3 Complementarity of trials and decision models 222
6.7 Optimizing screening schedules 225
6.7.1 Recursive approach 225
6.7.2 Variational approach 228.
Notes:
Includes bibliographical references (pages [233]-257) and index.
ISBN:
0471986089
OCLC:
47522464

The Penn Libraries is committed to describing library materials using current, accurate, and responsible language. If you discover outdated or inaccurate language, please fill out this feedback form to report it and suggest alternative language.

Find

Home Release notes

My Account

Shelf Request an item Bookmarks Fines and fees Settings

Guides

Using the Find catalog Using Articles+ Using your account