Transcript The pervasive influence of Homer R. Warner and The Biomedical

The pervasive influence of
Homer R. Warner
and
The Biomedical Informatics Department
at the University of Utah
Ten original contributions
Paul D. Clayton, PhD
Emeritus Professor & Former Chair
Medical Informatics at Columbia University
50th Anniversary Celebration – 18 April 2015
1958 Contract to computerize Index Medicus
1959 “Reasoning foundations of medical
diagnosis” Ledly, Lusted Science
1961 “A mathematical approach to medical
diagnosis. Application to congenital heart
disease.“ Warner HR, et. Al. JAMA
Mid 60’s Creation of academic departments:
US, France, Netherlands, Germany
Installation of first user applications
Signal processing
Systemized Nomenclature of Pathology
70’s Expert systems: DxPlain,Mycin,Internist,
Operational applications: Regenstrief,
HELP, DIOGENE, Kaiser,
CT, MRI, Image processing
80’s Standards (HL7, UMLS, LOINC,
SNOMED, ARDEN, GLIF, CEN TC 251)
Commercial providers of systems
Interoperability: FTP, e-mail, IAIMS
90’s WWW, (remote access) Natural
Language Processing, PACS
Database analytics
00’s Government incentives for use
1. Use of statistical probability in
diagnosis: Bayes’ theorem
• Homer’s 1961paper used real
patient data and observed
probabilities
2. Clinical System Architecture
Data entry
applications
Clinical
laboratory
results
Rule
evaluator
Data up
Retrieve
PTXT Dictionary
HELP sectors
Strato
Patient
Data
HCOM
Radiology
User
applications
Surgery
literature
Clinical Lab
Pathology
Interface
engine
Inference
engine
(CDSS)
PACS
Genetic tests
Anesthesia
vocabulary
Data
access
modules
Knowledge
base
Immunization
Insurance
...
Patient,
provider &
clinical
database
Data
warehouse
CIS Component Model
Data
Warehouse
MPI
Repository
CDSS
Knowledge
Base
Interface
Engine
Ancillary /
Departmental
Systems
Clinical
Applications
Data
Dictionary
RHIO
HIE
NwHIN
Other
Provider
Organizations
3. Clinical Database structure
• Entity, attribute, value (type 2)
• Defined columns for each entity
(type 1)
Even though technology has
changed dramatically during the
fifty years, The value of the clinical
data is unchanged
•
•
•
•
•
Punch cards vs voice input
Software architecture and tools
2k program overlays vs unlimited memory
Hardwired terminals vs hand-held WIFI
Film vs digital imaging
4. Dictionary which defines terms
stored in the database
PTXT (pointer to text)
UMLS
“The Metathesaurus forms the base of the UMLS and
comprises over 1 million biomedical concepts and 5
million concept names, all of which stem from the over
100 incorporated controlled vocabularies and
classification systems. Some examples of the
incorporated controlled vocabularies are ICD10, MeSH, SNOMED CT, DSM-IV, LOINC, WHO
Adverse Drug Reaction Terminology, UK Clinical
Terms, RxNorm, Gene Ontology,
and OMIM” (Wikipedia).
5. Representing clinical
Knowledge
• Declarative vs. Procedural
• HELP sectors
Arden Syntax
6.Evoking medical logic
• Data driven - Asking for data input
when the clinician is correct most of
the time in order to catch the 2% of
the time when there is a problem is
off-putting.
• ask mode (differential diagnosis)
• information content
• Decision analysis (help sectors for
probability and utility nodes)
Decision
analysis
Prostate cancer
p3
p2
p1
Impotent?
1-p3
Incontinent?
1-p2
Survive?
1-p1
Utility
Cost +QALY
Utility
Utility
Utility
Surgery?
yes
Radiation?
p4 metastases
Utility
1-p4
Utility
hormonal
p5
1-p5
hormonal
metastases
Utility
Utility
p6
Metastase
s
Utility
1-p6
Utility
7. Population analysis
STRATO
• find the population for whom
a particular Help sector logic
is true by applying that logic
to every patient in the
database
8. Genetic Markers
• Family pedigree Mark Skolnick 1974
• Finger print analysis
• Isolated genetic markers (BRCA etc.)_
9.Evaluation of the impact of
informatics
Gardner, Evans, Haug
• ICU outcomes
• Drug alerts
• Infectious disease occurrence and control
• Quality outcomes and subsequent
improvements
10. Student and faculty contributions
•
•
•
•
•
•
Arden
LOINC
HL-7
Genetic markers
UMLS
Leadership: Study section, AMIA, ACMI
academic departments, businesses,
Healthcare organizations