Clinical Decision Support and Beyond

Opis

Clinical Decision Support and Beyond: Progress and Opportunities in Knowledge-Enhanced Health and Healthcare, now in its third edition, discusses the underpinnings of effective, reliable, and easy-to-use clinical decision support systems at the point of care as a productive way of managing the flood of data, knowledge, and misinformation when providing patient care. Incorporating CDS into electronic health record systems has been underway for decades; however its complexities, costs, and user resistance have lagged its potential. Thus it is of utmost importance to understand the process in detail, to take full advantage of its capabilities. The book expands and updates the content of the previous edition, and discusses topics such as integration of CDS into workflow, context-driven anticipation of needs for CDS, new forms of CDS derived from data analytics, precision medicine, population health, integration of personal monitoring, and patient-facing CDS. In addition, it discusses population health management, public health CDS and CDS to help reduce health disparities. It is a valuable resource for clinicians, practitioners, students and members of medical and biomedical fields who are interested to learn more about the potential of clinical decision support to improve health and wellness and the quality of health care.

Presents an overview and details of the current state of the art and usefulness of clinical decision support, and how to utilize these capabilities
Explores the technological underpinnings for developing, managing, and sharing knowledge resources and deploying them as CDS or for other uses
Discusses the current drivers and opportunities that are expanding the prospects for use of knowledge to enhance health and healthcare

Spis treści

Front Cover 2
Clinical Decision Support and Beyond: Progress and Opportunities in Knowledge-Enhanced Health and Healthcare 5
Copyright 6
Contents 7
Contributors 13
Preface to the third edition, 2023 17
Preface to the second edition, 2014 21
Preface to the first edition, 2007 23
Section I: Goals, methodologies, and challenges for clinical decision support and beyond 29
- Chapter 1: Definition, purposes, and scope 31
  - 1.1. Introduction: CDS And beyond 31
    - 1.1.1. Definition and scope 36
    - 1.1.2. The evolution of CDS approaches and experiences 40
    - 1.1.3. CDS and beyond 43
    - 1.1.4. Scope and plan of this book 46

Front Cover 2
Clinical Decision Support and Beyond: Progress and Opportunities in Knowledge-Enhanced Health and Healthcare 5
Copyright 6
Contents 7
Contributors 13
Preface to the third edition, 2023 17
Preface to the second edition, 2014 21
Preface to the first edition, 2007 23
Section I: Goals, methodologies, and challenges for clinical decision support and beyond 29
- Chapter 1: Definition, purposes, and scope 31
  - 1.1. Introduction: CDS And beyond 31
    - 1.1.1. Definition and scope 36
    - 1.1.2. The evolution of CDS approaches and experiences 40
    - 1.1.3. CDS and beyond 43
    - 1.1.4. Scope and plan of this book 46
    - 1.1.5. Goals of the book 48
    - 1.1.6. What we do not cover 49
    - 1.1.7. Audiences 50
  - 1.2. CDS and the human 50
    - 1.2.2. Computer as out-of-touch meddler 51
    - 1.2.1. Computer as omniscient sage 51
    - 1.2.3. A more symbiotic view 52
    - 1.2.4. Limitations of the technology 53
    - 1.2.5. Considerations regarding human-computer interaction 54
  - 1.3. Design and structure of CDS 57
    - 1.3.1. Purpose 58
      - 1.3.1.1. Find information: Answer questions 59
      - 1.3.1.2. Make decisions-About diagnosis, test selection, choice of treatment, and prognosis 59
      - 1.3.1.3. Optimize process flow and workflow 62
      - 1.3.1.4. Monitor actions-guarding against errors, providing warnings, alerts, reminders, or feedback about performance 62
      - 1.3.1.5. Focus attention and enhance visualization 63
    - 1.3.2. Design of CDS: Components and interactions 64
      - 1.3.2.1. Execution engine - Responsible for implementing the decision method/model 65
      - 1.3.2.2. Knowledge content resources 66
      - 1.3.2.3. Information model 68
      - 1.3.2.4. Result specification 69
      - 1.3.2.5. Application environment 70
    - 1.3.3. Modes of interactions 72
      - 1.3.3.1. Locus of control 72
      - 1.3.3.2. Degree of assertiveness 73
      - 1.3.3.5. Interactivity 74
      - 1.3.3.3. Patient specificity 74
      - 1.3.3.4. Context 74
      - 1.3.3.7. Single vs. multi-party focus 75
      - 1.3.3.6. Degree of integration with clinical IT applications 75
      - 1.3.3.8. Intended user 76
      - 1.3.3.9. Explicit vs. indirect support 76
    - 1.4. Other considerations 76
    - References 77
  - Chapter 2: Clinical decision support methods 81
    - 2.1. Introduction 81
    - 2.2. Primary research methodologies that have been pursued and extended 81
      - 2.2.1. Information retrieval 83
        
        2.2.1.1. Taxonomy-based or ontology-based search 83
        
        2.2.1.3. Semi-automated or automated 85
        
        2.2.1.2. Free text search 85
        
        2.2.1.4. Question answering systems 86
      - 2.2.2. Evaluation of logical conditions 87
        
        2.2.2.1. Decision tables 87
        
        2.2.2.2. Venn diagrams 90
        
        2.2.2.3. Logical expressions 91
        
        Alerts and reminders 92
        
        2.2.2.4. Other logic models 93
      - 2.2.3. Probabilistic and data-driven classification or prediction 93
        
        2.2.3.1. Updating probabilities based on evidence 93
        
        2.2.3.2. Decision analysis 95
        
        2.2.3.5. Database prediction 97
        
        2.2.3.3. Bayesian belief networks 97
        
        2.2.3.4. Technology assessment 97
        
        2.2.3.6. Evidence-based medicine and meta-analysis 98
      - 2.2.4. Artificial intelligence 98
        
        2.2.4.1. Heuristic modeling and expert systems 99
        
        Rule-based systems 99
        
        Other heuristic modeling approaches 100
        
        2.2.4.2. Data mining and machine learning 102
      - 2.2.5. Calculations, algorithms, and multistep processes 103
        
        2.2.5.1. Interactive dialogue and structured data entry control 104
        
        2.2.5.2. Computer-based consultations 105
        
        2.2.5.3. Clinical practice guidelines 106
        
        2.2.5.4. Biomedical signal and image processing 107
      - 2.2.6. Associative groupings of elements 107
      - 2.2.7. Visualization 109
    - 2.3. Conclusion 110
    - References 111
  - Chapter 3: The journey to broad adoption 119
    - 3.1. The tale of a relationship 119
      - 3.1.1. A long infatuation - 1960s-1980s 121
        
        3.1.1.1. Simple and complex 121
        
        3.1.1.2. Evidence of usefulness 122
      - 3.1.2. A troubled courtship - 1980s-90s 123
        
        3.1.2.1. Recognizing that even simple CDS is a hard problem 124
        
        3.1.2.2. Implications for the technology of CDS 125
      - 3.1.3. Rekindled passions - Late 1990s-early 2000s 126
        
        3.1.3.1. Socioeconomic drivers 126
        
        3.1.3.2. Evolution of foundational technologies on which to develop CDS 127
      - 3.1.4. Getting the support of the relatives (stakeholders) - 2000s-present 128
      - 3.1.5. New parties to the relationship - Mid-2000s-present 130
        
        3.1.5.1. Addressing knowledge management and infrastructure 131
        
        3.1.5.2. Three intersecting and interacting lifecycles 132
        
        3.1.5.3. Broader infrastructure initiatives 136
      - 3.1.6. Building the foundations for a lasting relationship: New drivers for adoption - Late 2000s-present 137
        
        3.1.6.1. A more holistic view of health and healthcare 137
        
        3.1.6.2. New models of health care finance 139
        
        3.1.6.3. Science and technology drivers and enablers 140
      - 3.2. Where we go from here 144
      - References 144
    - Chapter 4: The role of quality measurement and reporting feedback as a driver for care improvement 151
      - 4.1. Introduction 151
      - 4.2. Quality measures and clinical decision support: Similarities and differences 153
      - 4.3. Creating a quality measure 153
      - 4.4. Constructing the quality measure equation 155
      - 4.5. Translating measure concepts into interoperable structures and definitions 159
      - 4.6. Identifying CDS interventions based on the quality measure 161
      - 4.7. A CDS rule component taxonomy 164
      - 4.8. The details about the CDS rules component taxonomy 165
      - 4.9. The CDS rules component taxonomy as a driver for quality measurement and care improvement 166
      - 4.10. Assuring the quality of a measure 168
      - 4.11. Driving care improvement 168
      - References 168
    - Chapter 5: International dimensions of clinical decision support systems 173
      - 5.1. Introduction 173
      - 5.2. Healthcare system organization in Australia and New Zealand 174
      - 5.3. Digital health context in Australia and New Zealand 175
        
        5.3.1. Existing health information infrastructures 175
        
        5.3.2. Clinical terminologies 176
        
        5.3.3. Data exchange 176
        
        5.3.4. Implications for patient safety governance of CDSSs 177
        
        5.3.4.1. Regulation of software as a medical device 177
        
        5.3.4.2. Clinical software standards 178
        
        5.3.4.3. Safety governance over CDSS deployment and use 178
        
        5.4. CDSSs in Australia and New Zealand 178
        
        5.4.2. Task-specific CDSSs in hospitals 179
        
        5.4.1. Task-specific CDSSs in general practice 179
        
        5.4.3. Process support CDSSs in general practice 181
        
        5.4.5. Section summary 183
        
        5.4.4. Process support CDSS in hospitals 183
        
        5.5. Healthcare system organization in Europe 184
        
        5.6. Digital health context in Europe 185
        
        5.6.1. Existing health information infrastructures in Europe 185
        
        5.6.3. Clinical terminologies in Europe 186
        
        5.6.2. Clinical software regulatory standards in Europe 186
        
        5.6.4. Data exchange in Europe 187
        
        5.7. CDSSs in Europe 187
        
        5.7.1. Task-specific CDSS implementation and use in Europe 187
        
        5.7.2. CDSSs across settings used in Europe 188
        
        5.7.4. Section summary 190
        
        5.7.3. National and cross-country CDSSs in Europe 190
        
        5.9. Healthcare system organization in low- and middle-income countries: Focus on East Africa and Haiti 191
        
        5.8. Emerging CDSS uses in high-income nations 191
        
        5.10. Health information systems in LMICs 193
        
        5.10.1. Mobile health (mHealth) systems 195
        
        5.10.2. Clinical software standards in LMICs 196
        
        5.10.3. Clinical terminologies in LMICs 196
        
        5.10.4. Standard vocabularies 196
        
        5.10.5. Data exchange in LMICs 197
        
        5.10.6. Open source software and global goods in LMICs 197
        
        5.11. CDSSs in low- and middle-income countries 198
        
        5.11.1. CDSSs delivered by mobile devices: mHealth 200
        
        5.11.2. Future trends 202
        
        5.12. Conclusions 203
        
        Acknowledgments 205
        
        References 205
      - Section II: Sources of knowledge for clinical decision support and beyond 217
        
        Chapter 6: Human-intensive techniques 219
        
        6.1. Introduction 219
        
        6.2. Theoretical basis for knowledge acquisition 224
        
        6.2.1. The nature of expertise 224
        
        6.2.2. Role of mental models 226
        
        6.2.3. Team-based decisions and shared knowledge 226
        
        6.3. Cognitive task analysis 227
        
        6.3.1. Knowledge elicitation (KE) methods 227
        
        6.3.1.2. Biases in logical and probabilistic reasoning 229
        
        6.3.1.1. Group techniques 229
        
        6.3.2. Data analysis methods 230
        
        6.3.2.1. Protocol and discourse analysis 230
        
        6.3.2.2. Concept analysis 231
        
        6.3.2.3. Verification and validation of knowledge acquisition 231
        
        6.3.2.4. Heuristic methods 232
        
        6.3.3. Representational methods 232
        
        6.3.4. Practical applications of representational approaches 234
        
        6.4. History and evolution of computer-based knowledge acquisition 235
        
        6.5. Example 237
        
        6.5.1. Case study: The Mayo Clinic knowledge management program 238
        
        6.6. Conclusions 238
        
        References 239
        
        Chapter 7: Data-driven approaches to generating knowledge: Machine learning, artificial intelligence, and predictive modeling 245
        
        7.1. Introduction 245
        
        7.2. Types of machine learning 248
        
        7.3. Frameworks for developing machine learning models in healthcare 249
        
        7.4. Learning from data 251
        
        7.4.1. Representative data selection 251
        
        7.4.2. Data bias 252
        
        7.4.3. Ethical and equitable use of data 252
        
        7.5. Overview of machine learning modeling methods 252
        
        7.5.1. Logistic regression 253
        
        7.5.2. Classification trees 256
        
        7.5.3. Support vector machines 259
        
        7.5.4. Artificial neural networks and deep learning 260
        
        7.6. Prediction models in medicine 263
        
        7.6.1. ICU severity of illness prediction 264
        
        7.6.2. Cardiovascular disease risk 265
        
        7.6.3. Pneumonia severity-of-illness index 266
        
        7.6.4. Machine learning for diagnostic prediction in medical imaging 267
        
        7.7. Machine learning CDS implementation 268
        
        7.8. Conclusion 272
        
        References 273
        
        Chapter 8: Modernizing evidence synthesis for evidence-based medicine 285
        
        8.1. Introduction 285
        
        8.2. Systematic reviews and meta-analysis: The premise and promise 286
        
        8.4. Steps of a systematic review 289
        
        8.4.1. Formulating the research question 290
        
        8.4.2. Searching the literature 291
        
        8.4.3. Screening 293
        
        8.4.4. Extracting data 293
        
        8.4.5. Conducting a meta-analysis 294
        
        8.4.5.1. Exploring heterogeneity 296
        
        8.4.5.2. Meta-analysis of complex datasets 298
        
        8.5. Accessing systematic reviews, meta-analyses, and field synopses 298
        
        8.6. Systematic reviews, meta-analyses, and the evidence ecosystem 300
        
        8.7. Conclusion 301
        
        References 303
        
        8.3. Uses of systematic reviews and meta-analyses 289
        
        Section III: The technology of clinical decision support and beyond 307
        
        Chapter 9: Decision rules and expressions 309
        
        9.1. Introduction 309
        
        9.2. Procedural knowledge 310
        
        9.3. Knowledge as production rules 312
        
        9.4. A hybrid approach for knowledge transfer: Arden Syntax 316
        
        9.5. Expression languages 319
        
        9.6. Standard data models for decision rules 321
        
        9.6.1. HL7 FHIR and data interoperability 321
        
        9.7. Toward further standardization: Quality measures and Health e-Decisions 322
        
        9.7.1. Clinical Quality Framework 323
        
        9.7.2. Clinical Quality Language 324
        
        9.7.3. Representation of clinical decision support rules 328
        
        9.7.5. Adapting clinical guidelines 333
        
        9.7.4. CDS Connect 333
        
        9.8. Future work 333
        
        References 334
        
        9.9. Conclusions 334
        
        Chapter 10: Guidelines and workflow models 337
        
        10.1. Introduction 337
        
        10.1.1. Increasing and standardizing quality of care via clinical practice guidelines 337
        
        10.1.2. Supporting management of routine medical actions via workflow systems 339
        
        10.1.3. The life-cycle of health care process and decision support systems 341
        
        10.2. Supporting knowledge acquisition 342
        
        10.2.1. The quality of narrative guidelines 342
        
        10.2.2. The types of knowledge contained in narrative guidelines 343
        
        10.2.3. From narrative to formal representations of guidelines 343
        
        10.2.4. From narrative to formal representations of clinical workflows (an example) 345
        
        10.3. Formal methods for modeling and specifying CIGs 349
        
        10.3.1. Task-network models for CIGs 350
        
        10.3.2. Other CIG modeling methods 355
        
        10.3.3. Workflow formalisms 356
        
        10.3.3.2. BPMN tools 358
        
        10.3.3.1. The building blocks of BPMN 358
        
        10.4. Integration of guidelines with workflow 359
        
        10.5. CIG and workflow verification and exception-handling 362
        
        10.6. CIG and careflow enactment tools 364
        
        10.7. Process mining and improvement 365
        
        10.8. CIG-based decision support for multimorbidity patients 366
        
        10.9. Support of patients as end-users via CIG and workflow models 367
        
        10.10. Discussion 369
        
        10.10.1. Successes 369
        
        10.10.2. Limitations 371
        
        10.10.3. Future research 371
        
        10.11. Recommended resources 372
        
        References 373
        
        Acknowledgment 373
        
        Chapter 11: Terminologies, ontologies and data models 377
        
        11.1. Introduction 377
        
        11.1.1. Aims of the chapter 377
        
        11.1.2. Referencing data in decision logic 379
        
        11.1.2.1. Facilitating portability among different information systems 379
        
        11.1.3. The need for coded data 380
        
        11.1.3.1. Advantages of coded data 380
        
        11.1.3.2. Enabling a ``Learning Health System´´ environment 381
        
        11.1.3.3. Coded data and ML/AI 381
        
        11.2. Introduction to the semantic spectrum: Ontologies, vocabularies/terminologies, and Data models 382
        
        11.2.1. The semantic spectrum 382
        
        11.2.1.1. Terminologies 382
        
        11.2.1.2. Ontologies 382
        
        11.2.1.3. Data models 383
        
        11.3. Formalisms and use cases 385
        
        11.3.1. Vocabulary formalisms 385
        
        11.3.2. Ontology formalisms 386
        
        11.3.3. Data model formalisms 388
        
        11.3.3.1. Name value pair paradigm in model formalisms 389
        
        11.3.4. Theory of combining the formalisms 389
        
        11.3.4.1. Semantic integrity in combining models with terminologies/ontologies 389
        
        11.3.4.2. The value of ontological principles in data modeling 390
        
        11.4. Implications 392
        
        11.4.1.1. Multiple name-value pair alternatives 392
        
        11.4.2. Pre- and post-coordination 392
        
        11.4.3. Placing information in the terminology model or the information model 394
        
        11.4.4. Need for explicit, specific models 396
        
        11.4.5. Iso-semantic models 396
        
        11.4.6. Protecting against changes in the terminology 397
        
        11.4.7. Context 398
        
        11.4.7.1. Dimensions of context 398
        
        11.4.7.2. Context and terminology services 399
        
        11.4.7.3. Context and individualized care 399
        
        11.5. Implementations of ontologies, terminologies, and models in CDS 399
        
        11.5.1. Implementation of terminology to support CDS interventions 399
        
        11.5.1.1. Authoring and browsing applications 400

Pokaż więcej

Progress and Opportunities in Knowledge-Enhanced Health and Healthcare

Opis

Spis treści

Recenzje (0)