Скачать презентацию Medical informatics Lecture 2 Formalising clinical data and Скачать презентацию Medical informatics Lecture 2 Formalising clinical data and

db710b7e2c7f6db5808f3f223e34168a.ppt

  • Количество слайдов: 42

Medical informatics Lecture 2 Formalising clinical data and medical knowledge, Clinical coding systems, Formal Medical informatics Lecture 2 Formalising clinical data and medical knowledge, Clinical coding systems, Formal knowledge representation … if time … formalising research results

Creating and using medical knowledge Understanding diseases and their treatment Develop and test treatments Creating and using medical knowledge Understanding diseases and their treatment Develop and test treatments Health Records Service delivery, performance assessment Ensure right Patients receive right intervention

… using medical knowledge Understanding diseases and their treatment Develop and test treatments Health … using medical knowledge Understanding diseases and their treatment Develop and test treatments Health Records Service delivery, performance assessment Ensure right Patients receive right intervention

Standardising clinical terms • Very difficult to use ordinary medical language in computer systems Standardising clinical terms • Very difficult to use ordinary medical language in computer systems – Extremely complex vocabulary. – Terms often vague and imprecise. – Same disease known by several names or expressions (synonymy). – A single term may have several meanings according to the context (polysemy). • Addressed by adopting formal coding and classification systems.

Formal coding and classification systems • Different systems use same code or term in Formal coding and classification systems • Different systems use same code or term in same way – Unique codes, precisely defined coding process • Benefits include abilities to – Share data between many systems – Gather data about diseases/treatments from many sources. – Deliver reminders, alerts and other information to clinicians based on standardised clinical patterns or situations – Identify eligible patients for recruitment into clinical trials based on well-defined criteria – Search professional literature based on standard queries • … and many other benefits

Coding systems • • International Classification of Diseases (ICD) Diagnosis Related Groups (DRGs) Standard Coding systems • • International Classification of Diseases (ICD) Diagnosis Related Groups (DRGs) Standard Nomenclature for Medicine (SNOMED) Logical Observation Identifiers Names and Codes (LOINC) • Medical Subject Headings (Me. SH) • Specialised coding systems – National Cancer Institute – Centre for Disease Control – …

WHO ICD • The International Classification of Diseases has been used since 1853 to WHO ICD • The International Classification of Diseases has been used since 1853 to classify diseases and other health problems recorded on many types of records, including death certificates and health records • In addition to enabling the storage and retrieval of diagnostic information for clinical, epidemiological and quality purposes, ICD also provides a basis for the compilation of national mortality and morbidity statistics by WHO Member States (e. g. AIDS, “swine flu”).

International Classification of Diseases I. Certain infectious and parasitic diseases II. Neoplasms III. Diseases International Classification of Diseases I. Certain infectious and parasitic diseases II. Neoplasms III. Diseases of the blood and blood-forming organs, immune mechanism IV. Endocrine, nutritional and metabolic diseases V. Mental and behavioural disorders VI. Diseases of the nervous system VII. Diseases of the eye and adnexa VIII. Diseases of the ear and mastoid process IX. Diseases of the circulatory system X. Diseases of the respiratory system XI. Diseases of the digestive system XII. Diseases of the skin and subcutaneous tissue XIII. Diseases of the musculoskeletal system and connective tissue XIV. Diseases of the genitourinary system XV. Pregnancy, childbirth and the puerperium XVI. Certain conditions originating in the perinatal period XVII. Congenital malformations, deformations and chromosomal abnormalities XVIII. Symptoms, signs and abnormal clinical and laboratory findings, not elsewhere classified XIX. Injury, poisoning and certain other consequences of external causes XX. External causes of morbidity and mortality XXI. Factors influencing health status and contact with health services XXII. Codes for special purposes

ICD family of disease and health related classifications Primary healthcare Information Support Other healthcare ICD family of disease and health related classifications Primary healthcare Information Support Other healthcare related classifications 3 -character core Speciality codes • Diagnoses • Symptoms • Abnormal Lab findings • Injuries and poisonings • External causes of morbidity and mortality • Factors influencing health status • oncology • dentistry • dermatology • psychology • neurology • obstetrics & gynaecology • rheumatology & orthopaedics • general medical practice International Nomenclature of Diseases

ICD-10 • Uses an alphanumeric code that indicates the location of the concept within ICD-10 • Uses an alphanumeric code that indicates the location of the concept within a disease hierarchy. • L 93 Lupus erythematosus. • Excludes exedens A 18. 4, vulgaris A 18. 4. . . • Use additional external cause code, if drug induced. – L 93. 0 Discoid lupus erythematosus – L 93. 1 Subcutae cutaneous lupus erythematosus

Diagnosis Related Groups • DRGs developed for relating the type of patients a hospital Diagnosis Related Groups • DRGs developed for relating the type of patients a hospital treats (“case mix”) to treatment costs • All discharged patients in US classified into a DRG – a limited, clinically coherent set of patient classes – based on age, sex, principal diagnosis, secondary diagnoses, surgical procedures, and discharge status • All patients are unique but groups of patients have common demographic, diagnostic and therapeutic attributes that determine resource needs.

DRGs 1 2 24 25 26 1 2 HYPERTENSIVE ENCEPHALOPATHY 23 NONTRAUMATIC STUPOR & DRGs 1 2 24 25 26 1 2 HYPERTENSIVE ENCEPHALOPATHY 23 NONTRAUMATIC STUPOR & COMA SEIZURE & HEADACHE AGE >17 WITH COMPLICATIONS, COMORBIDITIES (prior to 10 -1 -06) SEIZURE & HEADACHE AGE >17 WITHOUT COMPLICATIONS, COMORBIDITIES (prior to 10 -1 -06) SEIZURE & HEADACHE AGE 0 -17 TRAUMATIC STUPOR & COMA, COMA >1 HR TRAUMATIC STUPOR & COMA, COMA <1 HR AGE >17 WITH COMPLICATIONS, COMORBIDITIES 29 TRAUMATIC STUPOR & COMA, COMA <1 HR AGE >17 WITHOUT COMPLICATIONS, COMORBIDITIES 30 TRAUMATIC STUPOR & COMA, COMA <1 HR AGE 0 -17 31 CONCUSSION AGE >17 WITH COMPLICATIONS, COMORBIDITIES 32 CONCUSSION AGE >17 WITHOUT COMPLICATIONS, COMORBIDITIES

The Systematized NOmenclature of MEDicine (SNOMED) • Intended to be a general purpose, comprehensive The Systematized NOmenclature of MEDicine (SNOMED) • Intended to be a general purpose, comprehensive and computerinterpretable terminology • To represent and index “virtually all of the events found in the medical record”

SNOMED code for tuberculosis DE-14800 X-referencing Tuberculosis Bacterial infections E = Infections or parasitic SNOMED code for tuberculosis DE-14800 X-referencing Tuberculosis Bacterial infections E = Infections or parasitic diseases D = Disease or diagnosis X-ref e. g. living organism, morphology, function

Logical Observation Identifiers Names and Codes (LOINC) • A database and universal standard for Logical Observation Identifiers Names and Codes (LOINC) • A database and universal standard for identifying medical laboratory observations. • Applies universal code names and identifiers to medical terminology • For use in – gathering of clinical results (such as laboratory tests, clinical observations, outcomes management and research) – electronic data exchange – electronic health records

LOINC • LOINC currently includes over 58, 000 terms • A unique 6 -part LOINC • LOINC currently includes over 58, 000 terms • A unique 6 -part name is given to each test or observation. • Each database record includes six fields – Component- what is measured, evaluated, or observed – Kind of property - e. g. length, mass, volume, time stamp – Time interval over which observation or measurement made – System - or specimen type within which observation was made – Scale - quantitative, ordinal, nominal or narrative – Method procedure used to make measurement or observation

Problems with coding systems • Terms are subjective, often vague, imprecise • Codes often Problems with coding systems • Terms are subjective, often vague, imprecise • Codes often ad hoc (e. g. no systematic relationship between code and medical concepts and their uses) – Context dependent (e. g. “normal BP”) – Evolve over time • Mapping between systems difficult • Computers don’t “understand” them – we need to capture meaning rather than use ad hoc codes

Formalising medical concepts Understanding diseases and their treatment Develop and test treatments Health Records Formalising medical concepts Understanding diseases and their treatment Develop and test treatments Health Records Service delivery, performance assessment Ensure right Patients receive right intervention

Formalising medical concepts Models Rules breast cancer IS_A reproductive system and breast disorder Relationships Formalising medical concepts Models Rules breast cancer IS_A reproductive system and breast disorder Relationships Concepts DF-400 DB Symbols breast cyst breast lump breast cancer

REPRODUCTIVE SYSTEM AND BREAST DISORDERS IS_A WOMAN WITH POSSIBLE BREAST CANCER HAS-HEALTHCARE-PHENOMENON DISEASE OF REPRODUCTIVE SYSTEM AND BREAST DISORDERS IS_A WOMAN WITH POSSIBLE BREAST CANCER HAS-HEALTHCARE-PHENOMENON DISEASE OF THORAX HYPOTHESIS IS_A BREAST CANCER HYPOTHESIS IS-HYPOTHESIS-OF IS_A IS-RANGE-OF-DOMAIN-OF MALIGNANT NEOPLASM OF BREAST SPECIALIST IS_A IS-SPATIAL-PART-OF MALIGNANT NEOPLASM HAS-WE-STATE MALIGNANT BREAST FINDING BREAST DISEASE IS_A BREAST STRUCTURE IS-CONSEQUENCE-OF IS_A BREAST CANCER NEOPLASM IS_CREATIVE_RESULT_OF NEOPLASTIC PROCESS IS_A ADVANCED BREAST CANCER IS_A HAS-HEALTHCARE-PHENOMENON HAS-WE-STATE IS_A ADVANCED CANCER PATIENT IS_A PATIENT ADVANCED IS_A BREAST CANCER PATIENT Core Ontology breast cancer IS_A reproductive system and breast disorder

SNOMED CT All concepts (except the root) are the source of at least one SNOMED CT All concepts (except the root) are the source of at least one ‘ISA’(subtype) relationships Concept A concept is described in one or more descriptions Relationship For example the concept "headache (finding)" in SNOMED CT includes: • A concept. Id (25064002), Description • A set of descriptions ("headache", "pain in head", etc. ) • A set of relationships ("is a"="pain", "finding site"="head structure", etc. ).

Unified Medical Language System Links the major international terminologies into a common structure, providing Unified Medical Language System Links the major international terminologies into a common structure, providing a translation mechanism between them. Designed to aid in the development of systems that – retrieve and integrate electronic biomedical information from a variety of sources – permit linkage between disparate systems, including electronic patient records, bibliographic databases and decision support systems. “UMLS is the Rosetta Stone of international terminologies” A long term research goal is to enable computer systems to “understand” medical concepts.

UMLS semantic network • The UMLS Semantic Network allows for the semantic categorization of UMLS semantic network • The UMLS Semantic Network allows for the semantic categorization of a wide range of terminologies in multiple domains. • Major groupings of semantic types include – – – – organisms, anatomical structures, biologic function, chemicals, events, physical objects, and concepts or ideas. • Links between semantic types represent important relationships in the biomedical domain.

UMLS Meta-thesaurus • Uniformat for over 100 biomedical vocabularies and classifications • Organised by UMLS Meta-thesaurus • Uniformat for over 100 biomedical vocabularies and classifications • Organised by concept as a web rather than a tree, linking alternative names and views together and identifying useful relationships. – Components retain original structure. – Each concept has attributes that define its meaning (e. g. semantic types or categories to which it belongs, a definition). Clinical concept UMLS ICD-10 SNOMED CT Chronic ischaemic heart disease 448589 125. 9 14020 84537008

UMLS semantic network Links between semantic types represent important relationships in the biomedical domain. UMLS semantic network Links between semantic types represent important relationships in the biomedical domain. – Primary link: the isa link which establishes the hierarchy of types within the network – Secondary non-hierarchical relationships grouped into five major categories: physically related to, spatially related to, temporally related to, functionally related to, conceptually related to.

UMLS semantic network Content – For each semantic type: a unique identifier, a tree UMLS semantic network Content – For each semantic type: a unique identifier, a tree number indicating its position in an isa hierarchy, a definition, and its immediate parent and children. – For each relationship: a unique identifier, a tree number, a definition, and the set of semantic types that can plausibly be linked by this

Terms and ontologies http: //bioportal. bioontology. org/ Terms and ontologies http: //bioportal. bioontology. org/

Emulating clinical expertise • “Expert” systems offer “an engineering discipline that involves integrating [human] Emulating clinical expertise • “Expert” systems offer “an engineering discipline that involves integrating [human] knowledge into computer systems to solve complex problems normally requiring a high level of human expertise" – Successful early demonstrations were developed in chemistry, medicine, various fields of engineering (e. g. Banares-Alcantara’s work on design of chem eng. plant). • Key features distinguished expert systems from conventional software. – Explicit, declarative representation of knowledge: capture what an agent needs to know without assuming how that knowledge is to be used in any particular situation. – Domain-specific heuristics rather than general algorithms; said to resemble human expertise more closely than algorithmic methods.

Formalising medical concepts Models Lump and pre-menopausal implies possible breast cancer Rules breast cancer Formalising medical concepts Models Lump and pre-menopausal implies possible breast cancer Rules breast cancer is_a reproductive system and breast disorder Relationships Concepts DF-400 DB Symbols breast cyst breast lump breast cancer

CANCER SCENARIO IS_A REPRODUCTIVE SYSTEM AND BREAST DISORDERS IS-CCC-OF IS_A WOMAN WITH POSSIBLE BREAST CANCER SCENARIO IS_A REPRODUCTIVE SYSTEM AND BREAST DISORDERS IS-CCC-OF IS_A WOMAN WITH POSSIBLE BREAST CANCER HAS-HEALTHCARE-PHENOMENON IS_A DISEASE OF THORAX HYPOTHESIS IS_A BREAST CANCER HYPOTHESIS WOMAN WITH A REFERABLE BREAST LUMP IS-HYPOTHESIS-OF IS_A IS-RANGE-OF-DOMAIN-OF MALIGNANT NEOPLASM OF BREAST SPECIALIST IS_A IS-SPATIAL-PART-OF MALIGNANT NEOPLASM HAS-WE-STATE MALIGNANT BREAST FINDING BREAST DISEASE IS_A BREAST STRUCTURE IS-CONSEQUENCE-OF IS_A NEOPLASM BREAST CANCER STAGE I IS_CREATIVE_RESULT_OF IS_A NEOPLASTIC PROCESS ADVANCED BREAST CANCER IS_A HAS-HEALTHCARE-PHENOMENON ADVANCED HAS-WE-STATE IS_A ADVANCED CANCER PATIENT IS_A BREAST CANCER PATIENT Core Ontology Fragment of breast cancer domain knowledge: M Beveridge CRUK; W Ceusters Language and computing NV

Formalising medical concepts Models pre-menopausal woman has possible breast cancer (scenario) investigation of possible Formalising medical concepts Models pre-menopausal woman has possible breast cancer (scenario) investigation of possible breast cancer (process) Lump and pre-menopausal implies possible breast cancer Rules breast cancer is_a reproductive system and breast disorder Relationships Concepts DF-400 DB Symbols breast cyst breast lump breast cancer

Clinical research and evidence -based medicine NIH 1975 February 2010 Centre for Doctoral Training Clinical research and evidence -based medicine NIH 1975 February 2010 Centre for Doctoral Training

Levels of evidence February 2010 Centre for Doctoral Training Levels of evidence February 2010 Centre for Doctoral Training

Evidence-based medicine The practice of EBM has five steps: 1. Convert need for information Evidence-based medicine The practice of EBM has five steps: 1. Convert need for information about prevention, diagnosis, prognosis, therapy, causation etc. into an answerable question 2. Track down the best evidence to answer the question 3. Critically appraise the evidence for validity and applicability 4. Integrate the critical appraisal with our clinical expertise and our patient’s unique biology, values and circumstances 5. Evaluate our performance.

Medical research, clinical practice Understanding diseases and their treatment Develop and test treatments Health Medical research, clinical practice Understanding diseases and their treatment Develop and test treatments Health Records Service delivery, performance assessment Ensure right Patients receive right intervention

The clinical trial is “the most definitive tool for evaluation of the applicability of The clinical trial is “the most definitive tool for evaluation of the applicability of clinical research. ” Phase I Ascertain Maximum Tolerated Dose (MTD) Phase III Assess the effectiveness of a new intervention February 2010 Phase II Estimate effect and rate of adverse events Phase IV Long-term studies of licensed interventions Centre for Doctoral Training

February 2010 Centre for Doctoral Training February 2010 Centre for Doctoral Training

IT for evidence-based medicine Formalising research results “treatment comparison formulas” IT for evidence-based medicine Formalising research results “treatment comparison formulas”

IT for evidence-based medicine IT for evidence-based medicine

IT for evidence-based medicine IT for evidence-based medicine

Medical research, clinical practice Understanding diseases and their treatment Develop and test treatments Health Medical research, clinical practice Understanding diseases and their treatment Develop and test treatments Health Records Service delivery, performance assessment Lecture 3 Ensure right Patients receive right intervention