A BOOK ON STATISTICAL TOOLS FOR MEDICAL DECISION MAKING 

Medical decision making is evolving from an "art" to an evidence- and
model-based scientific discipline.  Physicians, patients, counselors,
policy makers and insurance companies are facing complex problems and
are addressing them based on an increasingly large amount of data. In
parallel, fast computing and recent progress in simulation-based
inference have led to a host of new and powerful statistical tools, many
of which have direct implications for medical decision making problems.
The potential for cross fertilization is substantial. 

The goal of the book I plan to write is to bridge the gap between
simulation-based Bayesian statistical methods and their use in medical
decision making, by providing a readable and detailed account of a
variety of case studies.  Most of the case studies will be based on
published papers of which I am one of the authors. In every case, the
material will be expanded compared to the original publications, with
several goals in mind: 1. To provide additional medical and/or policy
background and give details on the process of building the model and
identifying the appropriate statistical tools.  2. To provide further
background on the statistical techniques and to provide approaches to
simplified versions of the case study.  3. To provide a more detailed
account of the implementation than is typically possible in a journal
article.  4. To make the suite of applications consistent in terms of
notation and terminology and to increase the technical level as the book
proceeds.  In addition to the applications, the book will provide a self
contained and concise introduction to the methods. 



			Part 1. METHODS


[Ch 1]     Fundamentals of Bayesian inference. Practical use and
           interpretation of results.

[Ch 2]     Fundamentals of Utility theory. Use in individual and 
           societal decision making. Practical elicitation. 

[Ch 3]     Simulation in inference and decision-making. 
           Markov Chain Monte Carlo methods.  



			Part 2. CASES STUDIES 


[Ch 4]     Meta-analysis.

           Evaluating treatments for migraine headache. 


[Ch 5]     Elicitation of patient utilities.


[Ch 6]     Decision Trees and Influence Diagrams.

           Axillary node dissection in breast cancer 
        

[Ch 7]     Hierarchical models.

           Geographic variation of hospitalization costs after a
           stroke. (in progress)
        

[Ch 8]     Comprehensive decision models.

           The stroke prevention policy model. 
                

[Ch 9]     Cost-effectiveness and cost-utility analyses.

           Screening for breast cancer. 


[Ch 10]    Adaptive Decision Strategies
 
           Installation of heart defibrillators.