E-Book, Englisch, 1110 Seiten
Miner / Goldstein / Nisbet Practical Predictive Analytics and Decisioning Systems for Medicine
1. Auflage 2014
ISBN: 978-0-12-411640-5
Verlag: Elsevier Science & Techn.
Format: EPUB
Kopierschutz: 6 - ePub Watermark
Informatics Accuracy and Cost-Effectiveness for Healthcare Administration and Delivery Including Medical Research
E-Book, Englisch, 1110 Seiten
ISBN: 978-0-12-411640-5
Verlag: Elsevier Science & Techn.
Format: EPUB
Kopierschutz: 6 - ePub Watermark
Dr. Gary Miner PhD received a B.S. from Hamline University, St. Paul, MN, with biology, chemistry, and education majors; an M.S. in zoology and population genetics from the University of Wyoming; and a Ph.D. in biochemical genetics from the University of Kansas as the recipient of a NASA pre-doctoral fellowship. He pursued additional National Institutes of Health postdoctoral studies at the U of Minnesota and U of Iowa eventually becoming immersed in the study of affective disorders and Alzheimer's disease. In 1985, he and his wife, Dr. Linda Winters-Miner, founded the Familial Alzheimer's Disease Research Foundation, which became a leading force in organizing both local and international scientific meetings, bringing together all the leaders in the field of genetics of Alzheimer's from several countries, resulting in the first major book on the genetics of Alzheimer's disease. In the mid-1990s, Dr. Miner turned his data analysis interests to the business world, joining the team at StatSoft and deciding to specialize in data mining. He started developing what eventually became the Handbook of Statistical Analysis and Data Mining Applications (co-authored with Drs. Robert A. Nisbet and John Elder), which received the 2009 American Publishers Award for Professional and Scholarly Excellence (PROSE). Their follow-up collaboration, Practical Text Mining and Statistical Analysis for Non-structured Text Data Applications, also received a PROSE award in February of 2013. Gary was also co-author of 'Practical Predictive Analytics and Decisioning Systems for Medicine (Academic Press, 2015). Overall, Dr. Miner's career has focused on medicine and health issues, and the use of data analytics (statistics and predictive analytics) in analyzing medical data to decipher fact from fiction.Gary has also served as Merit Reviewer for PCORI (Patient Centered Outcomes Research Institute) that awards grants for predictive analytics research into the comparative effectiveness and heterogeneous treatment effects of medical interventions including drugs among different genetic groups of patients; additionally he teaches on-line classes in 'Introduction to Predictive Analytics', 'Text Analytics', 'Risk Analytics', and 'Healthcare Predictive Analytics' for the University of California-Irvine. Recently, until 'official retirement' 18 months ago, he spent most of his time in his primary role as Senior Analyst-Healthcare Applications Specialist for Dell | Information Management Group, Dell Software (through Dell's acquisition of StatSoft (www.StatSoft.com) in April 2014). Currently Gary is working on two new short popular books on 'Healthcare Solutions for the USA' and 'Patient-Doctor Genomics Stories'.
Autoren/Hrsg.
Weitere Infos & Material
1;Front Cover;1
2;Practical Predictive Analytics and Decisioning Systems for Medicine;4
3;Copyright Page;5
4;Contents;6
5;Foreword by Thomas H. Davenport;16
6;Foreword by James Taylor;18
7;Foreword by John Halamka;20
8;Preface;22
8.1;Modern Medicine: An Exercise in Prediction and Preparation;22
8.2;Wasted Costs in American Healthcare Systems;23
8.3;References;23
9;About the Authors;24
10;Acknowledgments;28
11;Guest Authors;30
12;Software Instructions;32
13;Introduction;34
13.1;Organization of This Book – Why We Did It This Way;36
13.1.1;Part 1;36
13.1.1.1;Part 1 – First Sub-Part;36
13.1.1.2;Part 1 – Second Sub-Part;36
13.1.2;Part 2;36
13.1.3;Part 3;36
14;Prologue to Part 1;38
14.1;Part 1 Historical Perspective and the Issues of Concern for Healthcare Delivery in the 21st Century;40
14.1.1;1 History of Predictive Analytics in Medicine and Health Care;42
14.1.1.1;Preamble;42
14.1.1.2;Background;42
14.1.1.3;Introduction;43
14.1.1.4;Part 1: Development of Bodies of Medical Knowledge;44
14.1.1.5;Earliest Medical Records in Ancient Cultures;44
14.1.1.6;Classification of Medical Practices IN Ancient and Modern Cultures;45
14.1.1.7;Medical Practice Documents in Major Ancient World Cultures of Europe and the Middle East;45
14.1.1.7.1;Egypt;45
14.1.1.7.2;Mesopotamia;46
14.1.1.7.3;Greece;47
14.1.1.7.3.1;Medicine in Pre-Classical Greece;47
14.1.1.7.3.2;Hippocrates and Classical Greece;48
14.1.1.7.4;Ancient Rome;49
14.1.1.7.4.1;Galen;49
14.1.1.7.5;Arabia;51
14.1.1.8;Summary of Royal Decrees Of Medical Documentation in Ancient Cultures;52
14.1.1.9;Effects of the Middle Ages on Medical Documentation;53
14.1.1.10;Rebirth of Interest in Medical Documentation During the Renaissance;53
14.1.1.10.1;The Printing Press;53
14.1.1.10.2;The Protestant Reformation;54
14.1.1.10.3;Erasmus;54
14.1.1.10.4;Human Anatomy;54
14.1.1.10.5;Andreas Vesalius (1514–1564);54
14.1.1.10.6;William Harvey (1578–1657);54
14.1.1.11;Medical Documentation Since the Enlightenment;55
14.1.1.11.1;Medical Case Documentation;55
14.1.1.11.2;The Development of the US National Library of Medicine;55
14.1.1.12;Part 2: Analytical and Decision Systems in Medicine and Health Care;56
14.1.1.13;Computers and Medical Databases;56
14.1.1.13.1;Medical Databases;57
14.1.1.13.1.1;Medical Literature Databases;57
14.1.1.14;Best Practice Guidelines;57
14.1.1.14.1;Guidelines of the American Academy of Neurology;58
14.1.1.15;Postscript;58
14.1.1.16;References;59
14.1.2;2 Why did We Write This Book?;60
14.1.2.1;Preamble;60
14.1.2.2;Introduction;60
14.1.2.3;Reason 1: Current Problems in Medical Research;61
14.1.2.3.1;Inaccuracies in Published Research Papers;61
14.1.2.3.2;Design Problems in Research Studies;62
14.1.2.3.3;A “Framework” for Determining Research Gaps;63
14.1.2.4;Reason 2: Practical Assistance is Needed to Insure Success for the New Initiatives and Accreditation Standards;64
14.1.2.4.1;The Joint Commission;64
14.1.2.4.2;Other Standards Organizations in Health Care;65
14.1.2.5;Reason 3: To Meet The Standards, Healthcare Organizations Need Practical Assistance and Tools With Implementing Lean Systems;66
14.1.2.5.1;Examples of Problems that Highlight the Need for “Lean” and Predictive Tools;66
14.1.2.5.1.1;Misdiagnoses;67
14.1.2.5.2;Six Sigma and Biomedical Informatics;68
14.1.2.6;Reason 4: Research into Technological/Organizational/Payment Changes Will be Necessary;70
14.1.2.6.1;Push Back in the Face of Change;70
14.1.2.6.1.1;Informed Consent and Transparency;70
14.1.2.6.1.2;Possible Loss of Innovation Due to Changes;70
14.1.2.6.2;Confusion in Coding and Payments Caused by Changes;71
14.1.2.6.3;Technology Difficulties;72
14.1.2.6.4;Organizational Culture;72
14.1.2.6.5;Population Studies Versus Patient-Focused Care;73
14.1.2.7;Reason 5: Practical Real World Examples are Needed that Bridge into A Phenomenal Future;73
14.1.2.7.1;Exploratory Statistics/Individualized Statistics/Predictive Statistics;73
14.1.2.7.2;Quality Medical Care Examples;74
14.1.2.7.3;Practical Predictive Analytics for the Lean Movement;74
14.1.2.7.3.1;Education and Tools;75
14.1.2.7.4;Back to the Future;75
14.1.2.8;Postscript;76
14.1.2.9;References;76
14.1.3;3 Biomedical Informatics;79
14.1.3.1;Preamble;79
14.1.3.2;The Rise of Predictive Analytics in Health Care;79
14.1.3.3;Moving From Reactive to Proactive Response in Health Care;80
14.1.3.4;Medicine and Big Data;80
14.1.3.5;An Approach to Predictive Analytics Projects;81
14.1.3.5.1;The Predictive Analytics Process in Health Care;82
14.1.3.5.2;Process Steps;82
14.1.3.5.2.1;Step 1: Problem Definition;82
14.1.3.5.2.2;Step 2: Identify Available Data Sources;83
14.1.3.5.2.3;Step 3: Formulate a Hypothesis;83
14.1.3.5.2.4;Step 4: Data Preprocessing;83
14.1.3.5.2.5;Step 5: Data Set Design;84
14.1.3.5.2.6;Step 6: Feature Selection;84
14.1.3.5.2.7;Step 7: Model Building;85
14.1.3.5.2.8;Step 8: Model Evaluation;85
14.1.3.5.2.9;Step 9: Model Implementation;85
14.1.3.5.2.10;Step 10: Validation of Clinical Utility;85
14.1.3.6;Meaningful Use;86
14.1.3.7;Translational Bioinformatics;86
14.1.3.8;Clinical Decision Support Systems;86
14.1.3.8.1;Hybrid CDSSs;87
14.1.3.9;Consumer Health Informatics;88
14.1.3.10;Direct-to-Consumer Genetic Testing;88
14.1.3.11;Use of Predictive Analytics to Avoid an Undesirable Future;89
14.1.3.12;Consumer Health Kiosks;89
14.1.3.13;Patient Monitoring Systems;89
14.1.3.14;Public Health Informatics;91
14.1.3.15;Medical Imaging;92
14.1.3.16;Clinical Research Informatics;93
14.1.3.17;Intelligent Search Engines;93
14.1.3.18;Personalized Medicine;93
14.1.3.19;Hospital Optimization;94
14.1.3.20;Challenges;94
14.1.3.21;Summary;95
14.1.3.22;Postscript;96
14.1.3.23;References;96
14.1.3.24;Further Reading;96
14.1.4;4 HIMSS and Organizations That Develop HIT Standards;97
14.1.4.1;Preamble;97
14.1.4.2;Introduction;97
14.1.4.2.1;Introduction to the Strategic Partners;98
14.1.4.3;Relationship Between ANSI, HIMSS, and ONC;99
14.1.4.4;Organizations Connected to or Influenced By HIMSS;99
14.1.4.5;Goals, Issues, and Ideals of HIMSS;99
14.1.4.6;ICD-10;100
14.1.4.7;HIMSS Attempts to Help;101
14.1.4.8;Standardization in Coding;102
14.1.4.9;Care Continuum Alliance (Another CCA) and Health Outcome Data;102
14.1.4.10;HIMSS Website;103
14.1.4.11;HIMSS Analytics;104
14.1.4.12;Progress of HIMSS;106
14.1.4.12.1;Compliance;106
14.1.4.12.2;Interoperability;106
14.1.4.13;Long-Range Problems and Opportunities;107
14.1.4.14;Some Questions;109
14.1.4.15;The Challenge;109
14.1.4.16;Postscript;109
14.1.4.17;References;110
14.1.5;5 Electronic Medical Records: Analytics’ Best Hope;111
14.1.5.1;Preamble;111
14.1.5.2;Introduction;111
14.1.5.3;What is an EMR?;112
14.1.5.4;A Bit (Of a “Byte”) of History …;113
14.1.5.5;Why aren’t We There Yet?;114
14.1.5.5.1;Cost;115
14.1.5.5.2;Usability;115
14.1.5.5.3;Disruption of Workflow;116
14.1.5.5.4;Lack of Interoperability;116
14.1.5.6;Ferraris and Country Roads;117
14.1.5.7;Postscript;122
14.1.5.8;References;122
14.1.5.9;Bibliography of Additional References on the Topic of Medical Records;123
14.1.6;6 Open-Source EMR and Decision Management Systems;124
14.1.6.1;Preamble;124
14.1.6.2;Introduction;124
14.1.6.3;Why Choose an Open-Source EMR Software Application?;125
14.1.6.4;Vista – The Veterans Administration System That Started it All;126
14.1.6.5;Five of the Best Open-Source EMR Systems for Medical Practices;126
14.1.6.5.1;The OSCAR EMR System;126
14.1.6.5.2;OpenEMR;127
14.1.6.5.2.1;OpenEmr Global Projects;128
14.1.6.5.2.1.1;Peace Corps;128
14.1.6.5.3;OpenMRS by Partners in Health;128
14.1.6.5.3.1;Mission of Partners in Health;128
14.1.6.5.3.2;Major OpenMRS Projects;129
14.1.6.5.3.2.1;Funding;130
14.1.6.5.3.3;PIH and OpenMRS Informatics Publications;130
14.1.6.5.4;Raxa Project;130
14.1.6.5.5;MOTECH in Ghana;130
14.1.6.6;Global Open-Source EMR Systems and the Future of Analytics;131
14.1.6.7;Postscript;131
14.1.6.8;References;132
14.1.7;7 Evidence-Based Medicine;133
14.1.7.1;Preamble;133
14.1.7.2;Introduction;133
14.1.7.3;Geodemographic Elements of Medical Treatment;134
14.1.7.4;How can we Define the Nature and Boundaries of EBM?;135
14.1.7.5;General Problems with EBM;135
14.1.7.6;Evidence-Based Medicine and Analytics;135
14.1.7.7;The Path to Evidence;136
14.1.7.8;What is a Randomized Controlled Trial?;137
14.1.7.9;If not Evidence Based, then What?;138
14.1.7.10;The EBM Process;139
14.1.7.11;Evidence at the Bedside;140
14.1.7.12;What do Patients Think?;140
14.1.7.13;Evidence-Based Medicine versus the Art of Medicine;141
14.1.7.14;Predictive Analytics and EBM;141
14.1.7.15;Postscript;141
14.1.7.16;References;141
14.1.8;8 ICD-10;143
14.1.8.1;Preamble;143
14.1.8.2;Introduction;143
14.1.8.3;Rise of the ICD;143
14.1.8.4;Why the ICD?;144
14.1.8.5;Elements Of ICD Documentation;144
14.1.8.6;The ICD Timetable;145
14.1.8.7;Changes Ahead for ICD-10 Users;145
14.1.8.8;Comparison of ICD-9 and ICD-10;146
14.1.8.8.1;Increased Ability to Describe and Justify Treatment;146
14.1.8.8.2;The ICD-10 Descriptive Language is Much Richer;146
14.1.8.8.3;Facilitation of Mortality and Morbidity Analyses;146
14.1.8.9;Implications of ICD-10 Changes;146
14.1.8.9.1;Greater Scalability and Extensibility Foster Information Sharing Among Institutions;146
14.1.8.9.2;More Specific Categories and Codes;147
14.1.8.9.3;Comparison of Codes;147
14.1.8.10;ICD-10 Codes in Practice;147
14.1.8.11;ICD-10 Changes in Terminology;148
14.1.8.12;Implementation Issues of Changing to ICD-10;148
14.1.8.13;What Lies Ahead for Payers and Providers?;149
14.1.8.13.1;For Providers;149
14.1.8.13.2;For Payers;149
14.1.8.14;Transition is a Joint Effort;149
14.1.8.15;Postscript;150
14.1.8.16;References;150
14.1.9;9 “Meaningful Use” – The New Buzzword in Medicine;153
14.1.9.1;Preamble;153
14.1.9.2;Introduction;153
14.1.9.3;Stage I of “Meaningful Use”;154
14.1.9.4;Meaningful Use Goals for Hospitals;154
14.1.9.4.1;The 14 Requirements (Hospitals Must Meet All of These);154
14.1.9.4.2;The 10 Choice Objectives (Hospitals Must Meet 5 of These);157
14.1.9.5;Meaningful Use Goals For Doctors;160
14.1.9.5.1;The 15 Requirements (Doctors Must Meet All of These);161
14.1.9.5.2;The 10 Additional Choice Objectives for Individual Physicians (5 of These Must Be Met to Achieve Compliance);163
14.1.9.6;Meaningful Use Requirements Of Stage I, Stage II, and Stage III;165
14.1.9.6.1;Requirements for Stage I of Meaningful Use;165
14.1.9.6.1.1;Medicare Providers;165
14.1.9.6.1.2;Medicaid Providers;166
14.1.9.6.1.3;Rcopia-MU Certified Modular EHR;166
14.1.9.7;Postscript;166
14.1.9.8;Bibliography;166
14.1.10;10 The Joint Commission;170
14.1.10.1;Preamble;170
14.1.10.2;History of the Joint Commission;170
14.1.10.3;The Joint Commission International;171
14.1.10.4;Joint Commission Accreditation;172
14.1.10.4.1;Preparing for a Survey;173
14.1.10.5;Other Regulatory Organizations;173
14.1.10.6;Joint Commission Standards;174
14.1.10.7;National Patient Safety Goals;175
14.1.10.8;Postscript;177
14.1.10.9;References;177
14.1.11;11 Root Cause Analysis, Six Sigma, and Overall Quality Control and Lean Concepts;180
14.1.11.1;Preamble;181
14.1.11.2;Introduction;181
14.1.11.3;Part 1: Six Sigma and Quality Control, Root Cause Analysis, and Leapfrog as they Developed During the 1990s and Early 2000s...;181
14.1.11.3.1;The Need For Quality: Medical Errors;181
14.1.11.3.1.1;Epidemiology of Medical Errors;181
14.1.11.3.1.2;Approaches to Error;181
14.1.11.4;Definitions;181
14.1.11.4.1;Statistical Process Control;181
14.1.11.4.2;Total Quality Management;182
14.1.11.4.3;Deming’s Principles;182
14.1.11.4.4;Six Sigma;182
14.1.11.4.5;Cost–Benefit Analysis;182
14.1.11.4.6;Pareto Efficiency;183
14.1.11.4.7;Kaldor-Hicks Efficiency;183
14.1.11.4.8;Examples of Errors;183
14.1.11.5;Methods to Improve Safety and Reduce Error;183
14.1.11.5.1;Root Cause Analysis;183
14.1.11.5.2;Ishikawa Diagram;183
14.1.11.5.3;Apollo Process;184
14.1.11.5.4;Systems for Ensuring Review;184
14.1.11.6;History of Quality in Health Care;184
14.1.11.6.1;Crossing the Quality Chasm: The IOM Health Care Quality Initiative;184
14.1.11.6.1.1;Phase One: 1996–1999;184
14.1.11.6.1.2;Phase Two: 1999–2001;184
14.1.11.6.1.3;Phase Three: 2001–Recent Present;185
14.1.11.6.2;Comprehensive Drug Safety;186
14.1.11.7;The Leapfrog Initiative;187
14.1.11.7.1;Organizational Goals of the Leapfrog Group;187
14.1.11.7.2;Why Leapfrog?;187
14.1.11.7.3;Leaps in Hospital Quality and Safety;188
14.1.11.7.4;Four Primary Criteria for Purchasing;188
14.1.11.7.5;Timeline;189
14.1.11.8;Part 2: Root Cause Analysis, Six Sigma and Quality Control, and Lean Concepts in Hospitals and Healthcare Facilities as The...;189
14.1.11.8.1;Part Outline;189
14.1.11.9;Six Sigma;189
14.1.11.10;Quality Control;190
14.1.11.10.1;Examples of Using Six Sigma in Health Care;190
14.1.11.11;Lean Concepts for Health Care: The Lean Hospital as a Methodology of Six Sigma;191
14.1.11.12;Root Cause Analysis;192
14.1.11.13;Part 3: Experiences of a Doctor who Implemented a Quality Control Department in a Hospital System During the 1990s – An Era...;195
14.1.11.14;Quality Improvement;195
14.1.11.14.1;General Introduction;195
14.1.11.14.2;Definition of Healthcare Quality;195
14.1.11.14.3;The Quality Department in a Hospital;196
14.1.11.14.4;Issues Discovered;197
14.1.11.15;Quality of Care Examples;199
14.1.11.15.1;Example 11.1;199
14.1.11.15.2;Example 11.2;200
14.1.11.16;Postscript;200
14.1.11.17;References;200
14.1.12;12 Lean Hospital Examples;202
14.1.12.1;Preamble;202
14.1.12.2;Introduction;202
14.1.12.3;Lean Kaizen Concepts;202
14.1.12.4;Henry Ford Hospitals;205
14.1.12.5;The Joint Commission Annual Report, 2013;205
14.1.12.5.1;Transparency Just Increased;205
14.1.12.6;Kaiser Permanente Managed Care Organization;206
14.1.12.7;Virginia Mason Hospital in Seattle;207
14.1.12.8;Examples of Lean Projects;208
14.1.12.8.1;Oncology: Infusion Therapy;208
14.1.12.8.2;Cardiology;209
14.1.12.8.3;Reducing Patient Falls;209
14.1.12.8.4;Reducing Catheter-Associated Urinary Tract Infections;210
14.1.12.8.5;Intravenous (IV) Laboratory Lean Project;210
14.1.12.8.6;Emergency Room Application of Lean;210
14.1.12.9;Summary;211
14.1.12.10;Postscript;211
14.1.12.11;References;211
14.1.13;13 Personalized Medicine;213
14.1.13.1;Preamble;214
14.1.13.2;What is Personalized Medicine?;214
14.1.13.2.1;P4 Medicine;215
14.1.13.2.2;P5 to P6 Medicine;215
14.1.13.3;Personalized Medicine, Genomics, and Pharmacogenomics;215
14.1.13.3.1;Differences Among Us;216
14.1.13.3.2;Differences Go Beyond Our Body and Into Our Environment;216
14.1.13.3.3;Changes from Birth to Death;216
14.1.13.3.4;Ancestry and Disease;217
14.1.13.3.5;It Is Not About Just Our Genome;217
14.1.13.4;Changing the Definition of Diseases;217
14.1.13.5;Systems Biology;218
14.1.13.6;Efficacy of Current Methods – Why We Need Personalized Medicine;219
14.1.13.7;Predictive Analytics in Personalized Medicine;220
14.1.13.8;The Future: Predictive and Prescriptive Medicine;220
14.1.13.8.1;Application of Predictive Analytics and Decisioning in Predictive and Prescriptive Medicine;220
14.1.13.9;The Diversity of Available Healthcare Data;222
14.1.13.9.1;Diversity of Data Types Available;222
14.1.13.9.1.1;1. Phenotypic Data;222
14.1.13.9.1.2;2. Clinical Information;223
14.1.13.9.1.3;3. Real-Time Physiological Data;223
14.1.13.9.1.4;4. Imaging Data;224
14.1.13.9.1.5;5. Genomic Data;224
14.1.13.9.1.5.1;DNA – The Centerpiece of Heredity and Bodily Differences;224
14.1.13.9.1.5.2;DNA Replication and Mutation;225
14.1.13.9.1.5.3;Somatic Mutations;226
14.1.13.9.1.5.4;Germline Mutations;226
14.1.13.9.1.5.5;The Personal Genome Project;227
14.1.13.9.1.5.6;The Electronic Medical Records and Genomics (eMERGE) Network;228
14.1.13.9.1.5.7;The Patient Centered Outcomes Research Institute (PCORI);228
14.1.13.9.1.6;6. Transcriptomics Data;228
14.1.13.9.1.7;7. Epigenomics Data;228
14.1.13.9.1.8;8. Proteomics Data;230
14.1.13.9.1.9;9. Glycomics Data;230
14.1.13.9.1.10;10. Metabolomics Data;230
14.1.13.9.1.11;11. Metagenomics Data;231
14.1.13.9.1.12;12. Nutrigenomics Data;232
14.1.13.9.1.13;13. Behavioral Measures Data;232
14.1.13.9.1.14;14. Socioeconomic Status Data;232
14.1.13.9.1.15;15. Personal Activity Monitoring Data;233
14.1.13.9.1.16;16. Climatological Data;234
14.1.13.9.1.17;17. Environmental Data;235
14.1.13.10;All the Other “Omics”;235
14.1.13.11;The Future;235
14.1.13.11.1;Challenges;237
14.1.13.11.1.1;Challenge 1;237
14.1.13.11.1.2;Challenge 2;237
14.1.13.11.1.3;Challenge 3;237
14.1.13.11.1.4;Challenge 4;238
14.1.13.11.1.5;Challenge 5;238
14.1.13.11.1.6;Challenge 6;238
14.1.13.11.1.7;Challenge 7;238
14.1.13.11.1.8;Challenge 8;238
14.1.13.11.1.9;Challenge 9;239
14.1.13.11.1.10;Challenge 10;239
14.1.13.12;Postscript;239
14.1.13.13;References;239
14.1.14;14 Patient-Directed Health Care;242
14.1.14.1;Preamble;242
14.1.14.2;The Empowered Patient;242
14.1.14.3;Patient Defined;243
14.1.14.4;Concept 1: Empowerment and Involvement – How can Patients be Empowered to Become More Involved with their Medical Care?;244
14.1.14.4.1;Patient Involvement;244
14.1.14.4.2;Hindrances to Patient Involvement;244
14.1.14.5;Concept 2: Coordination of Care and Communication;246
14.1.14.5.1;The Integrated Healthcare Delivery System Model;250
14.1.14.6;Concept 3: Consumerism in Health Care;250
14.1.14.7;Concept 4: Patient Payment Models;254
14.1.14.7.1;Burden of Health Care upon the Future;255
14.1.14.7.2;Mis-application of Treatment Increases Costs;255
14.1.14.7.3;Many Insurance Plans – Few Differences;255
14.1.14.8;Concept 5: Patient Education and Patient Self-Education and Decisions;257
14.1.14.8.1;Information Concerning Obesity;258
14.1.14.8.2;Patient Portals;260
14.1.14.8.3;Conclusion;260
14.1.14.9;Concept 6: Alternatives and New Models;260
14.1.14.9.1;Insurance Companies Going International;260
14.1.14.9.2;Alternative Screenings;261
14.1.14.9.2.1;An Alternative to Traditional Insurance;262
14.1.14.9.2.2;Doctors Striking out on their Own;262
14.1.14.9.3;Alternative Ways of Knowing about Ourselves;262
14.1.14.9.3.1;Genomic Predictions;262
14.1.14.9.3.2;Connectivity;265
14.1.14.9.3.2.1;Online Resources Connect Patients and Medical Personnel;265
14.1.14.9.3.2.2;Innovative Cleveland Clinic;266
14.1.14.9.3.2.3;Body Computing;266
14.1.14.9.3.2.4;Diagnosis Apps;267
14.1.14.10;Conclusion;268
14.1.14.11;Postscript;268
14.1.14.12;References;268
14.1.14.13;Further Reading;271
14.1.15;Prologue to Part 1, Chapter 15;274
14.1.16;15 Prediction in Medicine – The Data Mining Algorithms of Predictive Analytics;276
14.1.16.1;Preamble;276
14.1.16.2;Introduction;276
14.1.16.3;The Use of Simple Descriptive Statistics, Graphics, and Visual Data Mining in Predictive Analytics;277
14.1.16.3.1;The Insight of Simple Descriptive Statistics;277
14.1.16.3.2;Visual Data Mining;277
14.1.16.4;Predictive Modeling: Using Data to Predict Important Outcomes;279
14.1.16.4.1;The Difference Between Statistical Models and General Predictive Modeling;279
14.1.16.4.1.1;Traditional Statistical Analysis;279
14.1.16.4.1.2;Predictive Modeling Using General Approximators;280
14.1.16.4.2;The Algorithms of Predictive Modeling;280
14.1.16.4.2.1;k-Nearest Neighbor and Similar Methods;280
14.1.16.4.2.1.1;Benefits of the k-Nearest Neighbor Algorithm;281
14.1.16.4.2.1.2;Disadvantages of the k-Nearest Neighbor Algorithm;281
14.1.16.4.2.1.3;Extensions to the Approach;282
14.1.16.4.2.2;Recursive Partitioning Algorithms (Decision Trees);282
14.1.16.4.2.2.1;Example of a Simple Data Mining Application;282
14.1.16.4.2.2.2;Implementations;283
14.1.16.4.2.2.3;Strengths of Recursive Partitioning Algorithms;284
14.1.16.4.2.2.4;Weaknesses of Recursive Partitioning Algorithms;284
14.1.16.4.2.2.5;Early Stopping Techniques;284
14.1.16.4.2.3;Neural Networks;285
14.1.16.4.2.3.1;Benefits of Neural Networks Algorithms;285
14.1.16.4.2.3.2;Disadvantages of Neural Networks Algorithms;286
14.1.16.4.2.4;Model Ensembles;286
14.1.16.4.3;Choosing the Right Algorithm for the Right Analysis;286
14.1.16.4.3.1;Interpretable Models vs. “Black Box Models”;286
14.1.16.4.3.2;Interpolation and Extrapolation;287
14.1.16.5;Clustering: Identifying Clusters of Similar Cases, and Outliers;287
14.1.16.5.1;Clustering Algorithms;287
14.1.16.5.1.1;k-Means Clustering (and Expectation Maximization);287
14.1.16.5.1.1.1;Distances and Probabilities (Expectation Maximization Clustering);288
14.1.16.5.1.1.2;Differences Between the k-Means and EM Algorithms;289
14.1.16.5.1.1.3;Strengths of k-Means and EM Algorithms;289
14.1.16.5.1.1.4;Weaknesses of k-Means and EM Algorithms;289
14.1.16.5.1.2;Hierarchical or Tree Clustering;290
14.1.16.5.1.2.1;Strengths of Hierarchical Tree Clustering Algorithms;291
14.1.16.5.1.2.2;Weaknesses of Hierarchical Tree Clustering Algorithms;291
14.1.16.5.1.3;Kohonen Networks or Self-Organizing Feature Maps;291
14.1.16.5.1.3.1;Strengths and Weaknesses;292
14.1.16.6;Text Mining Algorithms;292
14.1.16.7;Dimension Reduction Techniques;292
14.1.16.7.1;Latent Semantic Indexing;292
14.1.16.7.2;Partial Least Squares;293
14.1.16.7.3;Feature Selection vs. Feature Extraction;294
14.1.16.7.3.1;Interaction Effects;294
14.1.16.8;Detecting the Interrelationships and Structure of Data Through Association and Link Analysis;295
14.1.16.8.1;The Support and Confidence Statistics;295
14.1.16.9;Summary;295
14.1.16.10;Postscript;296
14.1.16.11;References;296
15;Prologue to Part 2;298
15.1;Part 2 Practical Step-by-Step Tutorials and Case Studies;300
15.2;Guest Tutorial Authors;302
15.3;Tutorial A Case Study: Imputing Medical Specialty Using Data Mining Models;304
15.3.1;Bending the Curve;304
15.3.2;Identifying Cost-Efficient Physicians and Networks;305
15.3.2.1;The Episode Treatment Group Model;305
15.3.2.2;Building Episode Profiles for Physicians;306
15.3.2.3;Building Episode Cost Norms;307
15.3.2.4;Calculating Physician Cost-Efficiency;307
15.3.3;Why Physician Specialty is Important;307
15.3.4;Using ETG Data to Impute Specialty;308
15.3.4.1;Accuracy of the Criterion Variable;309
15.3.5;The Analysis Sample;309
15.3.6;Overview of the Data Mining Process;311
15.3.7;Data Mining Software;312
15.3.8;Data Mining Step by Step;312
15.3.8.1;Input Data;312
15.3.8.2;Redundant Variables;313
15.3.8.3;Selecting Key Predictors;314
15.3.8.4;Feature Selection;315
15.3.9;Testing the Data Mining Models;319
15.3.9.1;Running the SVM Model;319
15.3.9.1.1;Split-Sample Model Validation;322
15.3.10;Comparing the Performance of Different Models;323
15.3.10.1;SVM;324
15.3.10.2;MARS;324
15.3.10.3;C&RT;324
15.3.10.4;Random Forest;324
15.3.10.5;Boosted Trees;325
15.3.11;Additional Provider Information;326
15.3.12;Face Validity of the SVM Model Specialty Reassignment;326
15.3.13;Internal Medicine Reassignment;326
15.3.14;IM Remaining IM – ETG Frequency;326
15.3.15;IM Reclassified as Pulmonologist – ETG Frequency;327
15.3.16;IM Reclassified as Gastroenterologist – ETG Frequency;327
15.3.17;IM Reclassified as Cardiologist – ETG Frequency;327
15.3.18;IM Reclassified as Rheumatologist – ETG Frequency;327
15.3.19;Subspecialty Reassignment;328
15.3.20;Pulmonologist Reclassified as IM – ETG Frequency;329
15.3.21;Gastroenterologist Reclassified as IM – ETG Frequency;330
15.3.22;Cardiologist Reclassified as IM – ETG Frequency;330
15.3.23;Rheumatologist Reclassified as IM – ETG Frequency;330
15.3.23.1;Overall DM Model Assignment – Discussion;330
15.3.24;Models for General Practice/Family Practice;330
15.3.24.1;FP vs GP vs IM;331
15.3.25;Models for Pediatrics/General Surgery;331
15.3.26;General Comments;333
15.3.27;Testing Model Reliability;334
15.3.28;Postscript;334
15.4;Tutorial B Case Study: Using Association Rules to Investigate Characteristics of Hospital Readmissions;336
15.4.1;Objectives/Purpose;336
15.4.2;Common Readmission Conditions;337
15.4.3;Data Set;338
15.4.4;Association Rule Basics;340
15.4.5;Data Subsets;341
15.4.6;Generation of Association Rules;342
15.4.7;Adding Variables – Lift;345
15.4.8;Readmission Rules – 30 Days;346
15.4.9;Readmission Rules – 180 Days;350
15.4.10;Summary;351
15.4.11;References;352
15.5;Tutorial C Constructing Decision Trees for Medicare Claims Using R and Rattle;353
15.5.1;Objective;353
15.5.1.1;Variables;353
15.5.2;About Decision Trees;354
15.5.3;About Rattle;354
15.5.4;Data Preparation;354
15.5.5;Installing R;354
15.5.6;Installing Rattle;354
15.5.7;Starting Rattle after Installation;354
15.5.8;The Rattle Tab Bar;355
15.5.9;Importing the Tutorial Text File;355
15.5.10;Rattle Data Menu;357
15.5.10.1;Setting the Variable Roles;357
15.5.10.2;Exploring the Data;357
15.5.10.2.1;The Summary Analysis;357
15.5.10.2.2;The Distributions Analysis;358
15.5.10.2.3;Create Weighting Variable;359
15.5.10.2.4;Setting the Weight Variable Role;361
15.5.10.3;Running the Model;361
15.5.10.3.1;Saving the Model;363
15.5.10.3.2;Graphical View;363
15.5.10.3.3;Looking at a Specific Node;363
15.5.10.4;Evaluating the Model;364
15.5.10.5;Pruning the Tree;366
15.5.11;Example C2: Predicting Average Drug Cost For Medicare Part D;366
15.5.11.1;Transforming Average Drug Per Beneficiary to a Target Variable;367
15.5.11.1.1;Delete Observations with Missing Values;368
15.5.11.1.2;Bin the Variables;368
15.5.11.1.3;Exploring the Bins;369
15.5.11.2;Running the Conditional Model;370
15.5.11.3;Model Performance;374
15.5.11.3.1;Error Matrix;374
15.5.11.3.2;ROC Curve;375
15.5.12;The Rattle Log;375
15.5.13;Conclusion;376
15.5.14;Reference;377
15.5.15;Further Reading;377
15.6;Tutorial D Predictive and Prescriptive Analytics for Optimal Decisioning: Hospital Readmission Risk Mitigation;378
15.6.1;Overview;378
15.6.1.1;Statistical Data Analysis vs General Predictive (Pattern Recognition) Models;379
15.6.1.2;Example Data;379
15.6.2;The Goal;380
15.6.2.1;Step 1: Data Acquisition;380
15.6.2.2;Step 2: Feature Selection and Predictor Coding;381
15.6.2.2.1;Target Variable;381
15.6.2.2.2;Feature Selection;381
15.6.2.2.3;Predictor Coding, for more Interpretable Results;383
15.6.2.2.4;Weight-of-Evidence (WoE) Coding;384
15.6.2.3;Step 3: Predictive Modeling and Interpreting Results;384
15.6.2.3.1;Assessing the Quality of the Model: Lift Charts;388
15.6.2.3.2;Understanding the Implications of Results: What-If and Reason Scores;389
15.6.2.4;Step 4: Decision Management and Prescriptions;391
15.6.3;Conclusions;395
15.6.4;References;395
15.7;Tutorial E Obesity – Group: Predicting Medicine and Conditions That Achieved the Greatest Weight Loss in a Group of Obese/M...;396
15.7.1;Background;396
15.7.2;The Tutorial;396
15.7.2.1;Support Vector Machine (SVM);417
15.7.3;References;424
15.8;Tutorial F1 Obesity – Individual: Predicting Best Treatment for an Individual from Portal Data at a Clinic;425
15.8.1;Introduction;425
15.8.2;Background;426
15.8.3;The Exercise;428
15.8.3.1;Method 1: Weight of Evidence Using Individual Items for Beck Scores;428
15.8.3.2;Method 2: Weight of Evidence using Beck Scores Rather than Individual Items;454
15.8.3.3;Method 3: Using Weight (not Bifurcated) and Using Data Miner Recipe;454
15.8.4;References;482
15.9;Tutorial F2 Obesity – Individual: Automatic Binning of Continuous Variables and WoE to Produce a Better Model Than the “Han...;483
15.9.1;Introduction;483
15.9.2;The Exercise;483
15.10;Tutorial G Resiliency Study for First and Second Year Medical Residents;499
15.10.1;Introduction;499
15.10.1.1;The Instruments;499
15.10.1.1.1;Resilience Survey;500
15.10.1.1.2;TSCS2;501
15.10.1.1.3;Cattell 16PF Questionnaire;501
15.10.1.1.4;Myers & Briggs Type Indicator;502
15.10.2;Exercise G1: Predicting Year from Survey Questions;502
15.10.3;Exercise G2: Predicting Total Positive Resources and Total Negative Drain from 16PF and Myers & Briggs;521
15.10.4;Exercise G3: Predictive Analytics with Decisioning: Using Weight of Evidence;534
15.10.5;Appendix to Tutorial G;564
15.10.5.1;Special Instructions for Those Who Installed Version 12 Over an Older Version of STATISTICA;564
15.10.6;References;567
15.11;Tutorial H Medicare Enrollment Analysis Using Visual Data Mining;568
15.11.1;Introduction;568
15.11.2;Medicare Enrollment Data;568
15.11.3;Feature Selection and Root Cause Analysis;568
15.11.4;2D Mean Plot Analysis;574
15.11.5;Conclusion;580
15.12;Tutorial I Case Study: Detection of Stress-Induced Ischemia in Patients with Chest Pain After “Rule-Out ACS” Protocol;581
15.12.1;Background;581
15.12.2;Case Study;583
15.12.2.1;Methods;584
15.12.2.1.1;Study Design;584
15.12.2.1.2;Charts Selection;584
15.12.2.1.3;Variables;584
15.12.2.1.4;End Points;585
15.12.2.1.5;Statistical Analysis;585
15.12.2.2;Results;585
15.12.2.2.1;General Overview and Study Stages;585
15.12.2.2.2;Patient Characteristics and Flow in ACS “Rule Out” Algorithm (Step 1);586
15.12.2.2.3;Variables Screening (Step 2);589
15.12.2.2.4;Prediction Models Screening (Step 3);589
15.12.2.2.5;Model Evaluation (Step 4);590
15.12.2.2.5.1;Boosted Trees Model Performance;590
15.12.2.2.5.2;Neural Networks Model Performance;591
15.12.2.3;Discussion;591
15.12.3;References;592
15.13;Tutorial J1 Predicting Survival or Mortality for Patients with Disseminated Intravascular Coagulation (DIC) and/or Critical...;595
15.13.1;Disseminated Intravascular Coagulation in Critically Ill Patients;595
15.13.1.1;Diagnosis of DIC;596
15.13.1.2;Etiologies;598
15.13.1.3;Treatment of DIC;599
15.13.1.4;Proactivity in DIC: Effectiveness of Prophylactic Measures;599
15.13.2;Predictive Analytic Exercise;600
15.13.2.1;The Patients and Data;600
15.13.2.2;Examining the Data;600
15.13.2.3;Data Mining Recipe;607
15.13.2.4;SVM Model on the Fictionalized Data Set DIC1.Sta.;614
15.13.2.4.1;Feature Selection;614
15.13.2.5;SVM Model 1 – Using Data Mining Space;619
15.13.2.6;SVM Model 2 – Interactive Model;624
15.13.2.7;Text Mining;625
15.13.3;Conclusion;638
15.13.4;References;638
15.14;Tutorial J2 Decisioning for DIC;640
15.14.1;Introduction;640
15.14.2;Feature Selection;640
15.14.3;Weight of Evidence;646
15.14.4;Decisioning/Predictive Procedures;654
15.15;Tutorial K Predicting Allergy Symptoms;661
15.15.1;Introduction;661
15.15.2;Procedure;661
15.16;Tutorial L Exploring Discrete Database Networks of TriCare Health Data Using R and Shiny;672
15.16.1;Introduction;672
15.16.1.1;Objective of This Tutorial;672
15.16.1.1.1;The Data;673
15.16.1.2;About Database Structure and Network Analysis;674
15.16.1.3;About Shiny;674
15.16.1.4;Data Preparation;674
15.16.1.5;Installing R and RStudio;674
15.16.1.6;Installing Shiny and Supporting Packages;674
15.16.1.7;Starting the Tutorial Application;675
15.16.1.8;Features and Functionality of the Tutorial Application;676
15.16.2;Example I: Business-Driven Problem Generation – What do I Want to Ask of My Data?;678
15.16.2.1;The CRISP-DM and Other Analytics Processes;678
15.16.2.2;Exploring the Data;679
15.16.2.3;The Medical (Business) Problem;679
15.16.2.4;Inspect the Selected Database – “What Do We Have?”;680
15.16.2.5;Visual Exploration – “What Can We See?”;682
15.16.2.6;Database Relationships – “What Pieces of a Puzzle Do We Have?”;685
15.16.2.7;Variable Understanding and Connections – “How Can We Put the Pieces Together?”;686
15.16.2.7.1;Explore the Relationships Between ADDP CLAIMS and TDP CLAIMS;687
15.16.2.7.2;Explore the Relationships Between ADDP CLAIMS and DRF;689
15.16.2.7.3;Explore the Relationships Between ADDP CLAIMS and CDR CHEMISTRY;689
15.16.2.8;Iterating the Process – “Refining Our Understanding of the Data?”;691
15.16.2.9;Understanding What Is Missing in the Data;691
15.16.2.10;Conclusions – “Enough is Enough!”;692
15.16.3;Example 2: Data-Driven Problem Generation – “What Can My Data Inform Me of?”;693
15.16.3.1;Exploration on Your Own …;694
15.16.3.1.1;Business and Analytic Project Ideas;694
15.16.3.1.2;Ready, Set, Explore!;694
15.16.4;References;695
15.16.5;Further Reading;695
15.17;Tutorial M Schistosomiasis Data from WHO;696
15.17.1;Introduction;696
15.17.2;The Tutorial;697
15.17.2.1;Directions for the Tutorial;698
15.17.3;Part 1: Cleaning Data and Using Feature Selection as a Beginning Predictive Technique;698
15.17.4;Part 2: Examining the Original Data Using Statistica’s Data Health Check Module;714
15.17.5;Some Thoughts on Data Cleaning;733
15.17.6;References;733
15.18;Tutorial N The Poland Medical Bundle;734
15.18.1;Introduction;734
15.18.2;Data Verification;734
15.18.3;Missing Data Analysis;737
15.18.4;ROC Curves;738
15.18.5;Meta-Analysis and Meta-Regression;743
15.18.5.1;Study-Level Data Entry;743
15.18.5.2;Cumulative Meta-Analysis;747
15.18.5.3;Heterogeneity Analysis;747
15.18.5.4;Subgroup Analyses;748
15.18.5.5;Meta-Regression;749
15.18.5.6;Sensitivity Analysis;751
15.18.6;Logistic Regression Wizard;752
15.18.6.1;Simple Regression Analysis;755
15.18.6.2;Linearity of Predictors;756
15.18.6.3;Collinearity of Predictors;758
15.18.6.4;Interactions;758
15.18.6.5;Building a Multivariate Model;759
15.18.6.6;Model Evaluation;760
15.18.7;References;762
15.19;Tutorial O Medical Advice Acceptance Prediction;763
15.19.1;Introduction;763
15.19.2;Background;763
15.19.3;The Tutorial;764
15.20;Tutorial P Using Neural Network Analysis to Assist in Classifying Neuropsychological Data;782
15.20.1;Introduction;782
15.20.2;Examining the Data;782
15.20.3;Reference;793
15.21;Tutorial Q Developing Interactive Decision Trees Using Inpatient Claims (with SAS Enterprise Miner);794
15.21.1;About Decision Trees;794
15.21.2;About SAS© Enterprise Miner;795
15.21.2.1;Versions of Enterprise Miner;795
15.21.3;Data File Description;795
15.21.3.1;Target Variable;795
15.21.4;Creating a New Project;796
15.21.5;The Enterprise Miner Toolbar;796
15.21.6;Importing the Data File into SAS Using Enterprise Miner;798
15.21.7;Create a Data Source;799
15.21.8;Assigning Roles;799
15.21.8.1;Editing Variable Roles;799
15.21.9;Enterprise Miner Nodes;802
15.21.10;Reading a Data Source;802
15.21.11;Exploring the Data;803
15.21.12;Filtering the Data;805
15.21.13;Viewing Filter Results;808
15.21.14;Partitioning the Data;808
15.21.14.1;Add a “Data Partition” Node;808
15.21.14.2;Set Values on the Properties Pane;808
15.21.14.3;Run the Node;809
15.21.15;Decision Tree Modes;810
15.21.15.1;Node Splitting;810
15.21.15.2;Building the Interactive Tree;810
15.21.15.3;Changing the Statistics Displayed in Each Node;811
15.21.15.4;Splitting a Single Node;812
15.21.15.5;Terminal Nodes;816
15.21.15.5.1;Adjusting the Decision Tree Display;817
15.21.15.6;Training the Remaining Nodes;818
15.21.15.7;Collapsing/Expanding Branches;818
15.21.15.8;Copying a Second Version of a Tree You Just Created;820
15.21.15.9;Pruning the Tree;820
15.21.15.10;Evaluating Performance of the Model;822
15.21.15.10.1;Description of Model Outputs;822
15.21.15.10.1.1;Tree Map;822
15.21.15.10.1.2;Fit Statistics;824
15.21.15.10.1.3;Classification Chart;824
15.21.15.10.1.4;Cumulative Lift Chart;826
15.21.15.10.1.5;Leaf Statistics;826
15.21.15.10.1.6;Variable Importance;826
15.21.15.10.1.7;Resizing the Charts;826
15.21.15.11;Building an Automatic Binary Tree as an Alternative Model;827
15.21.15.11.1;Model Results;827
15.21.15.11.2;Model Comparison;828
15.21.15.11.3;Fit Statistics;830
15.21.16;Conclusion;831
15.21.17;Acknowledgments;831
15.22;Tutorial R Divining Healthcare Charges for Optimal Health Benefits Under the Affordable Care Act;832
15.22.1;Introduction;832
15.22.2;Pre-Tutorial Background on Original Data Set;833
15.22.3;Hospital Charge Analysis;835
15.22.3.1;Initial Data Exploration;837
15.22.3.2;Is There a Relationship Between Mean Charges and Status (Public or Private)?;837
15.22.3.3;Relationship of Charges to Location;838
15.22.4;Quality Matters;846
15.23;Tutorial S Availability of Hospital Beds for Newly Admitted Patients: The Impact of Environmental Services on Hospital Thro...;854
15.23.1;Introduction;854
15.23.2;Data Extraction;854
15.23.3;Running the Feature Selection for the EVS Throughput Tutorial Data Set;855
15.24;Tutorial T Predicting Vascular Thrombosis: Comparing Predictive Analytic Models and Building an Ensemble Model for “Best Pr...;869
15.24.1;Introduction;869
15.24.2;Tutorial;869
15.24.3;References;886
15.25;Tutorial U Predicting Breast Cancer Diagnosis Using Support Vector Machines;887
15.25.1;Introduction;887
15.25.1.1;Data Description;887
15.25.2;Data Analysis and Exploration;888
15.25.2.1;Feature Selection and Root Cause Analysis (Using Chi-square Method (Default)/p Value Method);888
15.25.3;Modeling Using Support Vector Machine with Deployment;894
15.25.4;Rapid Deployment, Cross-Validating, and Predicting on Different Data;899
15.25.5;Summary;902
15.25.6;Data Set Locations;902
15.26;Tutorial V Heart Disease: Evaluating Variables That Might Have an Effect on Cholesterol Level (Using Recode of Variables Fu...;903
15.26.1;Aim;903
15.26.2;Tutorial Steps;903
15.27;Tutorial W Blood Pressure Predictive Factors;913
15.27.1;Background;913
15.27.2;Data Review;914
15.27.3;Data Preparation;915
15.27.3.1;Discontinuity;915
15.27.3.2;Date Abstractions;915
15.27.3.3;Data Consolidation;917
15.27.3.4;Rolling Sum of Running Data;918
15.27.3.5;Climate Data;919
15.27.4;Data Importation;919
15.27.4.1;Step 1;919
15.27.4.2;Step 2;920
15.27.5;Variable Typing;923
15.27.6;Pattern Discovery;924
15.27.6.1;Descriptive Statistics;924
15.27.6.2;Feature Selection;927
15.27.7;Conclusion;932
15.28;Tutorial X Gene Search and the Related Risk Estimates: A Statistical Analysis of Prostate Cancer Data;933
15.28.1;Background and the Benchmark Data;933
15.28.2;Visualization (I): Categorized Histograms and Matrix Plots;934
15.28.3;The Benjamini-Hochberg FDR (False Discovery Rate);936
15.28.4;Prescreening and Dimension Reduction;939
15.28.5;Lasso, Adaptive Lasso, and Elastic Net;941
15.28.6;Hold-Out Data and Over-Fitting;942
15.28.7;Penalized Support Vector Machines;942
15.28.8;Conflicting Results from the Tree Methods;943
15.28.9;Visualization (II): Linear vs. Non-Linear Models;943
15.28.10;The Benjamini-Hochberg FDR and Non-Parametric Tests;945
15.28.11;Hybrid Methods;947
15.28.11.1;Sequential Hybrids;947
15.28.11.2;Eclectic Hybrids;948
15.28.12;Visualization (III): Seeing Can Be Deceiving;948
15.28.13;Biomarkers and Visualization (IV);950
15.28.14;Concluding Remarks and the Limitations of Statistical Analysis of Gene Data;952
15.28.15;Appendices;953
15.28.15.1;Appendix A: Efron 2010, Top 10 Genes;953
15.28.15.2;Appendix B: The R Code for penalizedSVM, svm.fs Function Usage;954
15.28.15.3;Appendix C: Stochastic Gradient Boosting;955
15.28.15.4;Appendix D: The R Code for Kolmogorov-Smirnov, Mann-Whitney, and Anderson-Darling Tests;956
15.28.16;References;956
15.29;Tutorial Y Ovarian Cancer Prediction via Proteomic Mass Spectrometry;958
15.29.1;Background and the Data;958
15.29.2;Data Preprocessing (I) and Dynamic Binning;958
15.29.3;Prediction Accuracies of Competing Models;962
15.29.4;False Positive, False Negative, and the Roc Index;964
15.29.5;Probability of Cancer;966
15.29.6;Limitation of Mass Spectrometry and Data Preprocessing (II);967
15.29.7;Variable Selection and Gene Search;969
15.29.8;Appendices A–C;971
15.29.8.1;Appendix A: The Top 10 Variables Selected by SAS-PLS;971
15.29.8.2;Appendix B: The Top 10 Variables Selected by Benjamini-Hochberg FDR;971
15.29.8.3;Appendix C: R Code for the Penalized SVM;972
15.29.9;References;975
15.30;Tutorial Z Influence of Stent Vendor Representatives in the Catheterization Lab;976
15.30.1;Introduction and Review of the Literature;976
15.30.2;Definition of Terms;977
15.30.3;Method of Gathering Data;977
15.30.4;Exploratory Analysis;978
15.30.5;Drug Eluting Stent Output;990
15.30.6;Conclusion;1000
15.30.7;References;1000
16;Prologue to Part 3;1002
16.1;Part 3 Practical Solutions and Advanced Topics in Administration and Delivery of Health Care Including Practical Predictive...;1004
16.1.1;16 Predictive Analytics in Nursing Informatics;1006
16.1.1.1;Preamble;1006
16.1.1.2;Introduction;1006
16.1.1.3;Nursing Informatics;1007
16.1.1.3.1;Patient Education;1007
16.1.1.3.2;Supporting Nurses’ Work;1008
16.1.1.3.3;Predicting the Patient’s Future Development;1008
16.1.1.3.4;Patient Monitoring Data;1008
16.1.1.3.5;Home Nursing and Nursing Homes;1008
16.1.1.3.6;Telemedicine;1009
16.1.1.3.7;Triaging Patients;1009
16.1.1.3.8;Preventing Inpatient Morbidity;1009
16.1.1.3.9;Patient Comfort and Satisfaction;1009
16.1.1.3.10;Staffing;1010
16.1.1.3.11;Patient Hand-Offs;1010
16.1.1.3.12;Approach to Projects;1010
16.1.1.4;Postscript;1010
16.1.1.5;References;1010
16.1.2;17 The Predictive Potential of Connected Digital Health;1012
16.1.2.1;Preamble;1013
16.1.2.2;Why Don’t Clinicians Embrace Digital Consumer Connections?;1013
16.1.2.2.1;1. What Do I Do with the Data?;1014
16.1.2.2.1.1;No Training;1014
16.1.2.2.1.2;No Time;1014
16.1.2.2.1.3;No Money;1014
16.1.2.2.1.4;No Space;1014
16.1.2.2.2;2. Who Says That the Data Are Valuable?;1014
16.1.2.2.2.1;Useful Data;1014
16.1.2.2.2.2;Scientific Method;1015
16.1.2.2.2.3;Long Timeframe;1015
16.1.2.2.2.4;Peer Review;1015
16.1.2.2.2.5;Accelerated Learning;1015
16.1.2.2.3;3. What New Liabilities Emerge from Precision and Probabilistic Medicine?;1015
16.1.2.2.3.1;Litigation Risk;1015
16.1.2.2.3.2;Data Quality;1015
16.1.2.2.3.3;Data Vigilance;1016
16.1.2.2.3.4;Probabilistic Predictions;1016
16.1.2.2.3.5;Transparency Risks;1016
16.1.2.2.3.6;Creation of Hypochondriacs;1016
16.1.2.2.4;4. How Cumbersome and Difficult Are New Data Collection Solutions?;1016
16.1.2.2.4.1;Lack of Solutions;1016
16.1.2.2.4.2;Mimic of the Drug Industry Business Model;1016
16.1.2.2.4.3;Vetting Applications;1017
16.1.2.2.4.4;Outsourced Monitoring;1017
16.1.2.2.4.5;Business Models;1017
16.1.2.2.5;5. How Do the Devices and Apps Integrate and Interoperate?;1017
16.1.2.2.5.1;Closed Systems;1017
16.1.2.2.5.2;Open Systems;1018
16.1.2.2.5.3;Workflow Integration;1018
16.1.2.2.5.4;Data Freedom;1018
16.1.2.2.6;6. How Do You Maintain Privacy and Security with Mobile Consumer Engagement?;1018
16.1.2.2.6.1;Some Safeguards are in Place;1018
16.1.2.2.6.2;Risks are Low;1018
16.1.2.2.6.3;Follow the Money;1018
16.1.2.2.7;7. How and When Will Clinicians Get Paid for Participating in Mobile Health?;1019
16.1.2.2.7.1;It’s More Than Money;1019
16.1.2.2.7.2;Money Matters;1019
16.1.2.3;Promise and Problems of Shifting to Mobile Health Technology;1019
16.1.2.3.1;Shifting to Mobile Health Technology May Not Lead to Additional Revenue;1019
16.1.2.3.2;Mobile Disruption;1019
16.1.2.3.3;Consumer Health;1020
16.1.2.3.4;Outcomes Matter;1020
16.1.2.4;What Can We Learn from the VA About the Potential of Predictions?;1020
16.1.2.5;What Can We Learn from Financial Services Regarding Digital Transformation?;1022
16.1.2.5.1;The Rise of Mobile Financial Services;1022
16.1.2.5.2;How Do These Five Insights from the Digitization of the Financial Services Industry Inform Our Views About the Digitization...;1023
16.1.2.5.2.1;Timing of Digital Transformation;1023
16.1.2.5.2.2;Local Customer Pull;1023
16.1.2.5.2.3;Emerging Dominance;1024
16.1.2.5.2.4;Healthcare Digital Complexity;1024
16.1.2.5.2.5;Digital Data Capture;1024
16.1.2.6;Summary and Recommendations;1025
16.1.2.7;Postscript;1025
16.1.2.8;References;1025
16.1.3;18 Healthcare Fraud;1026
16.1.3.1;Preamble;1026
16.1.3.2;Introduction;1026
16.1.3.3;Leakage Due to Fraud;1027
16.1.3.4;Definition of Fraud in the Healthcare Context;1027
16.1.3.4.1;Fraud Perpetrated by a Provider;1028
16.1.3.4.1.1;Billing for Services Not Provided;1028
16.1.3.4.1.2;Billing for Low Value or Unwarranted Services, and Incorrect Reporting of Diagnoses or Procedures to Maximize Payments;1028
16.1.3.4.1.3;Treatment Upcoding and Code Unbundling;1028
16.1.3.4.1.4;Misrepresentation of Dates or Descriptions of Services to Create Additional Eligible Charges;1028
16.1.3.4.2;Fraud Perpetrated by a Patient Subscriber;1028
16.1.3.4.2.1;Selling or Lending Covered Healthcare Identity to Others;1028
16.1.3.4.2.2;Fraudulent Enrollment in a HealthCare Plan;1028
16.1.3.4.2.3;Schemes to Fraudulently Obtain Prescription Medication;1028
16.1.3.4.3;Fraud Perpetrated by Third Parties;1028
16.1.3.4.4;Fraud Perpetrated by Agents/Brokers;1028
16.1.3.5;Statutes and Regulations Intended to Prevent, Detect, and Prosecute Fraud;1029
16.1.3.6;Major Agencies Involved in Healthcare Anti-Fraud Efforts;1029
16.1.3.7;Challenges That Face Anti-Fraud Efforts;1029
16.1.3.7.1;Traditional Challenges;1030
16.1.3.7.1.1;Non-Certainty of Detection;1030
16.1.3.7.1.2;Overlapping Patchwork of Enforcement;1030
16.1.3.7.1.3;Budget;1030
16.1.3.7.1.4;Size of Industry;1030
16.1.3.7.1.5;Organized Crime Efforts;1030
16.1.3.7.1.6;Privacy Statutes and Data Sharing;1030
16.1.3.7.2;Emerging Challenges;1030
16.1.3.7.2.1;Big Data;1030
16.1.3.8;Traditional Means of Detection;1031
16.1.3.8.1;Limitations of Traditional Means of Detection;1031
16.1.3.8.1.1;Reactive Detection;1031
16.1.3.8.1.2;Ad hoc Investigations;1031
16.1.3.8.1.3;Traditional Detection Systems;1031
16.1.3.8.1.4;Inability to Effectively Triage Investigations;1031
16.1.3.8.1.5;Inability to Incorporate Unstructured Information;1031
16.1.3.9;The Emergence of Big Data in Healthcare Investigations;1031
16.1.3.9.1;ACA Anti-Fraud Provisions;1031
16.1.3.10;Analytical Anti-Fraud Approaches;1032
16.1.3.10.1;Anomaly Detection;1032
16.1.3.10.2;Text Analytics;1032
16.1.3.10.2.1;Supervised Learning Techniques and Predictive Analytics;1032
16.1.3.10.3;Link Analysis;1032
16.1.3.10.4;Combined Analytical Techniques;1032
16.1.3.11;The Future of Healthcare Anti-Fraud Efforts;1032
16.1.3.12;Anti-Fraud Organizations;1033
16.1.3.13;Postscript;1033
16.1.3.14;References;1033
16.1.4;19 Challenges for Healthcare Administration and Delivery: Integrating Predictive and Prescriptive Modeling into Personalize...;1035
16.1.4.1;Preamble;1035
16.1.4.2;Challenges;1035
16.1.4.3;Postscript;1037
16.1.4.4;References;1038
16.1.5;20 Challenges of Medical Research for the Remainder of the 21st Century;1039
16.1.5.1;Preamble;1039
16.1.5.2;Challenges;1039
16.1.5.3;Postscript;1040
16.1.6;21 Introduction to the Cornerstone Chapters of this Book, Chapters 22–25: The “Three Processes” – Quality Control, Predicti...;1041
16.1.6.1;Preamble;1041
16.1.6.2;Introduction;1041
16.1.6.3;Traditional Statistics vs Data Mining vs Predictive Analytics;1042
16.1.6.4;Postscript;1044
16.1.7;22 The Nature of Insight from Data and Implications for Automated Decisioning: Predictive and Prescriptive Models, Decision...;1045
16.1.7.1;Preamble;1045
16.1.7.2;Overview;1045
16.1.7.3;The Nature of Insight and Expertise;1046
16.1.7.3.1;Procedural and Declarative Knowledge;1046
16.1.7.3.2;Non-Conscious Acquisition of Knowledge;1046
16.1.7.3.2.1;The Nature of the “Non-Conscious”;1047
16.1.7.3.3;Conclusion: Expertise and the Application of Pattern Recognition Methods;1047
16.1.7.4;Statistical Analysis vs Pattern Recognition;1047
16.1.7.4.1;Fitting a priori Models;1048
16.1.7.4.2;Pattern Recognition: Data are the Model;1048
16.1.7.4.2.1;Data are the Model;1048
16.1.7.4.2.2;Pattern Recognition Via General Approximators;1048
16.1.7.4.3;Pattern Recognition and Declarative Knowledge: Interpretability of Results;1050
16.1.7.4.3.1;Statistical Models, and Reason Scores;1050
16.1.7.4.3.2;Pattern Recognition Algorithms and Reason Scores;1050
16.1.7.4.3.3;What-if, and Reason Scores as Derivatives;1050
16.1.7.5;Predictive Modeling and Prescriptive Models;1051
16.1.7.5.1;Rules, Conditional Scoring Logic, Action Plans;1052
16.1.7.5.2;An Example System: The STATISTICA Enterprise Decisioning Platform®;1052
16.1.7.5.2.1;STATISTICA Enterprise Metadata Repository;1053
16.1.7.5.2.1.1;Metadata, and Control;1053
16.1.7.5.2.1.2;Prediction Models, and Rules;1053
16.1.7.5.2.1.3;STATISTICA Enterprise Server for Batch Scoring;1054
16.1.7.5.2.1.4;STATISTICA Live Score Server for Real-Time Scoring;1054
16.1.7.5.2.1.5;Monitoring and Alerting Server;1054
16.1.7.5.2.1.6;Document Management System for Version Control of Models, Model Management;1054
16.1.7.6;Summary;1055
16.1.7.7;Postscript;1055
16.1.7.8;References;1055
16.1.8;23 Platform for Data Integration and Analysis, and Publishing Medical Knowledge as Done in a Large Hospital;1056
16.1.8.1;Preamble;1056
16.1.8.2;Introduction;1056
16.1.8.3;Functions and Applications of the Platform;1057
16.1.8.4;Platform Components and Architecture;1057
16.1.8.4.1;Data Warehouse and OLAP;1058
16.1.8.4.2;Data Collection Module;1058
16.1.8.4.3;Medical Research Environment – Overview;1059
16.1.8.4.3.1;MRE – Data Filtering;1060
16.1.8.4.3.2;MRE – Column Selection;1060
16.1.8.4.3.3;MRE – Spreadsheet Layout;1060
16.1.8.4.3.4;MRE – Spreadsheet Generator;1062
16.1.8.4.3.5;MRE – Statistics;1062
16.1.8.4.4;Management Portal;1062
16.1.8.4.5;Reports for NHF Contract Monitoring and Clearance;1064
16.1.8.4.6;Optimizer;1066
16.1.8.5;Conclusions;1066
16.1.8.6;Postscript;1066
16.1.8.7;References;1066
16.1.9;24 Decisioning Systems (Platforms) Coupled With Predictive Analytics in a Real Hospital Setting – A Model for the World;1067
16.1.9.1;Preamble;1067
16.1.9.2;Introduction;1067
16.1.9.3;Setting the Stage for a Decisioning Platform;1068
16.1.9.3.1;Getting Support from Information Technology and Hospital Leadership;1068
16.1.9.3.2;Creating an Analytical Culture (Or, Have You Ever Tried to Tell a Surgeon He’s Doing Things Wrong?);1069
16.1.9.3.3;Defining the Outcomes Targets;1070
16.1.9.3.4;Defining the Clinical Decisions;1070
16.1.9.3.5;Define the Resources That Need to be Managed;1070
16.1.9.3.6;Determine What Data You Have Access to;1071
16.1.9.3.6.1;External Data;1071
16.1.9.3.6.2;EHR Data;1071
16.1.9.4;Deploying the Decision Management System;1071
16.1.9.4.1;Decision Management System Tools;1071
16.1.9.4.2;Decision Management Process;1072
16.1.9.4.3;Decision Management System Workflow Example;1072
16.1.9.5;Conclusion;1073
16.1.9.6;Postscript;1074
16.1.9.7;References;1074
16.1.10;25 IBM Watson for Clinical Decision Support;1075
16.1.10.1;Preamble;1075
16.1.10.2;Introduction;1075
16.1.10.3;Personalized Health Care and Clinical Decision Support;1075
16.1.10.4;IBM Watson and Medical Decision-Making;1076
16.1.10.5;Postscript;1077
16.1.10.6;References;1077
16.1.11;26 21st Century Health Care and Wellness: Getting the Health Care Delivery System That Meets Global Needs;1078
16.1.11.1;Introduction;1078
16.1.11.1.1;Overview;1079
16.1.11.2;Background and Need for Change;1079
16.1.11.3;Learning Objectives;1080
16.1.11.4;Trends Impacting Healthcare Industries;1080
16.1.11.5;Existing and Emerging Healthcare Organizations;1081
16.1.11.6;Health Start-Ups and Established Technology Firms Contributing to Health Care;1082
16.1.11.6.1;IBM Watson;1082
16.1.11.6.2;New Technology and 21st Century Health Care: Health Start-Up Firms;1083
16.1.11.6.3;Building the Star Trek Tricorder;1083
16.1.11.6.4;Wearable Computers for Doctors;1085
16.1.11.6.5;Explorys;1085
16.1.11.7;Technology Trends That Impact Health and Wellness;1086
16.1.11.7.1;Current Trends Outside Healthcare Facilities;1086
16.1.11.7.1.1;The eCAALYX Example;1086
16.1.11.8;Trends and Expectations for the Future of Health It and Analytics;1086
16.1.11.8.1;The Next 4 Years – by-2018 Predictions;1087
16.1.11.8.2;The Next 9 Years – by-2023 Predictions;1087
16.1.11.9;Conclusions and Summary of Important Concepts Presented in This Book;1088
16.1.11.9.1;Technology for the Elderly;1088
16.1.11.9.2;Technology for Rural Areas;1089
16.1.11.9.3;Final Concluding Statements;1089
16.1.11.10;References;1090
16.1.11.11;Bibliography;1091
17;Index;1092
About the Authors
Linda A. Winters-Miner, Ph.D., earned her bachelor’s and master’s degrees at University of Kansas, her doctorate at the University of Minnesota, and completed post-doctoral studies in psychiatric epidemiology at the University of Iowa. While she, with her husband Gary Miner, raised their children, Becky and Matt, she spent most of her career as an educator, in teacher education and statistics and research design. She spent nearly two years as a site coordinator for a major (Coxnex) drug trial. For 23 years, Miner directed academic programs for Southern Nazarene University–Tulsa. Her program direction included three undergraduate programs in business and psychology and three graduate programs in management, business administration, and health care administration. She has authored or co-authored numerous articles and books including with Gary and others, and the first book concerning the genetics of Alzheimer's: Linda authored some of the tutorials in the first two predictive analytic books published in 2009 and 2012 by Elsevier. At present, she teaches both undergraduate statistics & research at SNU-Tulsa, teaches statistics and predictive analytics for the IHI Family Practice Medical Residency program in Tulsa, and also teaches predictive analytics online, including healthcare predictive analytics, for both the University of California–Irvine and University of California–San Diego.
Pat Bolding, M.D., F.A.A.F.P. is a practicing board certified family physician. He has used an EMR (Electronic Medical Record) since his residency training in the mid 1980 s – at the time it was the “pioneering” Technicon Medical Information System. Later, as the CEO of a large family practice group (which also hosted a 30-resident training program), he led the selection and implementation of several EMR systems, beginning with the text-based Medic Autochart then Misys EMR and finally the A4-Healthmatics system. In 2007, he joined a multi-specialty group practice/integrated delivery system where he serves on the EMR committee that oversaw the implementation of the NextGen ambulatory EMR. More recently he was a member of the search committee that chose the Epic system to replace NextGen. He is a frequent speaker on health/medical topics and has a special interest in evidence-based medicine. He is an adjunct faculty member of Southern Nazarene University, teaching in the Health Care MBA program.
Thomas Hill, Ph.D., received his Vordiplom in psychology from Kiel University in Germany and earned an M.S. in industrial psychology and a Ph.D. in psychology and quantitative methods from the University of Kansas. He was associate professor (and then research professor) at the University of Tulsa from 1984 to 2009, where he taught data analysis and data mining courses. He also has been vice president for Research and Development and then the Analytic Solutions section at StatSoft Inc., where he has been involved for over 20 years in the development of data analysis, data and text mining algorithms, and the delivery of analytic solutions. Dr. Hill joined Dell through Dell’s acquisition of StatSoft (www.StatSoft.com) in April 2014, and he is currently the Executive Director for Analytics at Dell’s Information Management Group. Dr. Hill has received numerous academic grants and awards from the National Science Foundation, the National Institute of Health, the Center for Innovation Management, the Electric Power Research Institute, and other institutions. He has completed diverse consulting projects with companies from practically all industries and has worked with the leading financial services, insurance, manufacturing, pharmaceutical, retailing, and other companies in the United States and internationally on identifying and refining effective data mining and predictive modeling solutions for diverse applications. Dr. Hill has published widely on innovative applications for data mining and predictive analytics. He is the author (with Paul Lewicki, 2005) of , the (a popular on-line resource on statistics and data mining), a co-author of (2012); he is also a contributing author to the popular (2009).
Bob Nisbet, Ph.D. was trained initially in Ecology and Ecosystems Analysis. He has over 30 years experience in complex systems analysis and modeling, most recently as a Researcher (University of California, Santa Barbara). In business, he pioneered the design and development of configurable data mining applications for retail sales forecasting, and Churn, Propensity-to-buy, and Customer Acquisition in Telecommunications, Insurance, Banking, and Credit industries. In addition to data mining, he has expertise in data warehousing technology for Extract, Transform, and Load (ETL) operations, Business Intelligence reporting, and data quality analyses. He is lead author of the (Elsevier Academic Press, 2009), and a co-author of (Elsevier Academic Press, 2012). Currently, he serves as an Instructor in the University of California, Irvine Predictive Analytics Certification Program, teaches online courses in Effective Data preparation, and Introduction to Predictive Analytics.
Mitchell Goldstein, M.D., attended the University of Miami’s Honor Program in Medical Education under an Isaac B. Singer full tuition scholarship, completed his pediatric residency training at the University of California, Los Angeles, and finished his Neonatal Perinatal Medicine training at the University of California, Irvine in 1994. Dr. Goldstein is board certified in both Pediatrics and Neonatal Perinatal Medicine. He is an Associate Professor of Pediatrics at Loma Linda University Children’s Hospital and Emeritus medical director of the Neonatal Intensive Care Unit at Citrus Valley in West Covina, CA. He has been in clinical practice for 20 years. At the various places he has worked, Dr. Goldstein has become fluent in a multitude of EMRs including EPIC, Cerner, and Meditech. As a member of the Department Deputies Users Group at Loma Linda University Hospital, Dr. Goldstein participates in an ongoing EMR improvement process. Dr. Goldstein is a past president of the Perinatal Advisory Council, Legislation, Advocacy and Consultation (PACLAC) as well as a past president of the National Perinatal Association (NPA). Dr. Goldstein is the twice recipient of the annual Jack Haven Emerson Award presented to the physician with the most promising study involving innovative pulmonary research and the 2013 recipient of the National Perinatal Association Stanley Graven lifetime achievement award presented for his ongoing commitment to the advancement of neonatal and perinatal health issues. He is the editor of PACLAC’s Neonatal Guidelines of Care as well as the Principal author of both the National Perinatal Association’s 2011 Best Practice Checklist – Oxygen Management for Preterm Infants and Respiratory Syncytial Virus (RSV) Prophylaxis 2012 Guidelines. Dr. Goldstein serves on the editorial board of the Journal of Perinatology as well as Neonatology Today, has represented the NPA to the American Academy of Pediatrics (AAP) perinatal section, and is a moderator of NICU-NET, a neonatal listserv. He is an executive board member and is on the nominations committee for the Section on Advances in Therapeutics & Technology (SOATT) of the AAP. Dr. Goldstein chaired the NPA National Conferences in 2004, 2008 and 2011 and continues to be active in conference planning as the CME Continuing Medical Education (CME) chair for PACLAC. His research interests include the development of non-invasive monitoring techniques, evaluation of signal propagation during high frequency ventilation, and data mining techniques for improving quality of care. Dr. Goldstein has also been a vocal advocate for RSV prophylaxis and “right” sizing technology for the needs of neonates. Dr. Goldstein’s recent publications have included “Critical Complex Congenital Heart Disease (CCHD)” which was dual published in and , the “Late Preterm Guidelines of Care” published in the , and “How Do We COPE with CPOE” published in .
Joseph M. Hilbe, J.D., Ph.D., is an emeritus professor at the University of Hawaii, an adjunct professor of statistics at Arizona State University, and a Solar System Ambassador with NASA/Jet Propulsion Laboratory, Caltech. An elected Fellow of the American Statistical Association and elected member of the International Statistical Institute, Dr. Hilbe is currently President of the International Astrostatistics Association, is a full member of the American Astronomical Society, and Chairs the Statistics in Sports section of the American Statistical Association. He has authored fifteen books in statistical modeling, and over 200 book chapters, encyclopedia entries, journal articles, and published statistical software, and is currently on the editorial board of seven academic journals. During the 1990 s, Dr. Hilbe was on the founding executive...
