OpenEHR Foundation

1
OpenEHR Foundation

• Tutorial
• Peter Schloeffel
• Dipak Kalra
• Tom Beale
• David Lloyd
• Sam Heard

2
Tutorial themes

• Scene setting the challenge
• OpenEHR design principles
• Why do we need two models?
• Outline of the OpenEHR reference model,
  comparison with ENV13606
• Discussion of the Reference Model and Archetype
  Model
• Clinical archetype examples

3
Information needs in health care

• Challenges and changes
• Shift of care from specialist centres to
  community settings
• Increasing complexity of health care provision
• Requirement to deliver evidence-based care
• Increasingly distributed and mobile clinical
  workforce
• Critical reliance upon comprehensive patient
  records
• Overwhelming growth of medical knowledge
• Growth of consumerism and patient participation
  in health care
• Increasing concerns about the confidentiality of
  patient records

4
Challenges facing todays health record systems

• The need to record more data
• The need to analyse more data
• The need to share more data

5
The need to record more healthcare data

• patients receive more health care
• the information is more complex
• a richer variety of examinations, investigations,
  treatments
• a greater use of multi-media technologies
• the information is more important
• in case of future litigation
• to justify the use of healthcare resources

6
1999 US Dept. of Commerce study

• health care is the fourth most information
  reliant industry sector after
• telephone telegraph
• radio television
• security brokerage

7
UK Audit Commission Survey (1995)

• 15 of hospital resources are spent gathering
  information
• Up to 25 of doctors and nurses time is spent
  collecting and using information

8
Scale of the problem

• University Hospital of Heidelberg 1700 beds
• creates about 400,000 new medical records per
  year
• containing 6.3 million pages
• requiring 1.7 km of storage
• (growing at the rate of 1500m per annum)
• Physicians create over 250,000 reports and 20,000
  procedure reports each year
• service departments create around a million
  results
• Reinhard, Ohr, et al Haux 1998

9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
The need to analyse more healthcare data

• to observe trends and patterns within the
  historical record of one patient
• to enable the use of clinical guidelines and
  decision support tools evidence based health
  care
• to perform clinical audit
• to inform management and commissioning decisions
• to support epidemiology, research and teaching

13
1995 Audit Commission Report on Patient Health
Records

• 36 of casenotes not immediately available
• Multiple records for same patient in 75 of
  hospitals
• 30 of history sheets inadequate
• 20 of prescriptions illegible
• 40 of handwritten discharge medication sheets
  illegible

14
Importance of data quality and its availability

• The US Institute of Medicine report "To Err is
  Human" has estimated that 100,000 US citizens die
  each year through medical errors
• Medical errors may rank as the eighth leading
  cause of death in the US, and contribute 4
  (37.6 billion) to the cost of US healthcare
  Anderson 2000

15
Requirements of evidence based medicine

• the learned literature has doubled every 10-15
  years over the past 300 years
• in the field of biomedicine 20,000 journals and
  17,000 books are produced every year

16
Growing numbers of concepts

• Around half of the concepts believed to exist in
  the world (500,000) are in the medical domain
  Baud, Lovis, et al. 1998

17
Growing numbers of concepts

• Growth of the Columbia Presbyterian MED
• 1994 32,000 terms Cimino 1994
• 2000 60,000 concepts, 208,000 synonyms, 84,000
  hierarchic relations, 114,000 other semantic
  relations Cimino et al 2000
• Size of SNOMED
• 1979 45,000 terms
• 2000 157,000 terms
• MeSH contains 19,000 main subject headings
• UMLS Metathesaurus 730,000 concepts, 1.5 million
  terms

18
Growth of SNOMED (from www.SNOMED.org)
19
The need to share more healthcare data

• with other clinicians in the same team
• clinical firms, practice partnerships or nursing
  shifts
• with other healthcare professions
• doctors, nurses, physiotherapists, midwives,
  dieticians...
• with other disciplines
• a diabetic patient may also be under
  ophthalmology, nephrology, orthopaedics,
  chiropody, wheelchair clinic..
• with other institutions
• with patients and their families

20
Scale of the communications challenge

• England and Wales
• 200 health authorities
• 2,500 hospitals
• 400 Primary Care Groups
• 10,000 GP practices (35,000 GPs)
• 800,000 staff

At least 40 million shared care communications
per annum
21
Changing this...
22
into this
23
Problems with existing clinical systems

• workstations are too slow and clumsy for
  real-time use (or just not available!)
• structured templates and term sets are resented
  by clinicians
• patient encounters rarely follow a consistent
  pattern
• duplication of data entry still occurs
• existing paper records are usually retained
• the lack of agreed standards for the electronic
  transfer of records between systems

24
The Electronic Health Record
Principal challenges

• the diversity, complexity and evolutionary nature
  of clinical data
• making it difficult to capture, store and
  communicate
• the diversity and heterogeneity of current health
  record systems
• making it difficult to share the data between
  computer systems or to transfer data between
  sites
• the ethical and legal requirements of good
  clinical practice

25
Challenge of representing health information

• Large number of evolving concepts
• Comprehensive models
• difficult to agree
• difficult to maintain
• Systems designers need to respond to specialist
  needs, but to remain interoperable
• It is difficult to standardise the information
  requirements of one health domain
• without becoing too prescriptive
• whilst permitting the future evolution of health
  care

26
Delivering the EHR
Bringing together the patient record information
scattered across health care sites

• to provide seamless and secure sharing of EHRs
• derived from diverse clinical databases and
  record systems
• within large healthcare organisations
• across regions between primary and secondary care
• between countries
• implemented through Federated Health Record
  Server(s)

27
Combining feeder system data
Individual interfaces -gt combinatorial explosion
28
Handling a request for information
FHR services
Get...
clinical data X from patient Y
Server routes the request for X as defined in
the Archetype Object Dictionary
Feeder systems networked to FHR server
Client clinical workstation
29
Returning information in response to a request
FHR Services
Server returns a set of Record Components of
class X for patient Y
Feeder systems networked to FHR server
Client clinical workstation
30
Record federation
distributed access to local and remote feeder
systems
Client views the data requested
Local and Remote Feeder Systems
FHR Services
Get...
Client clinical workstation
Decommissioned or off-line feeder
31
Getting the right level of abstraction

• HIPPA legislation includes over 1000 fields
• "The solution is to use more abstract models with
  fewer, but more expressive, objects. The patient
  information model has to be simplified and
  clarified and a uniform and correct level of
  abstraction must be found... For example, items
  such as birth weight or haemoglobin concentration
  should not be found included as named attributes
  of the model. The specific clinical entities
  should be represented in a concept/vocabulary
  data base that is separate from the data model."
• McDonald, Overhage, et al. 1998

32
Reference List

• Anderson J.G., Social Research Institute, Purdue
  University, West Lafayette, Indiana, USA.
  Evaluating clinical information systems a step
  towards reducing medical errors. MD Computing.
  May 2000-30 Jun 2000 17(3)21-3.
• Baud R.H., Lovis C., Rassinoux A.M., and Scherrer
  J.R.Alternative ways for knowledge collection,
  indexing and robust language retrieval. Methods
  of Information in Medicine. Nov 1998
  37(4-5)315-26.
• Cimino J.J., Clayton P.D., Hripcsak G., and
  Johnson S.B. Knowledge-based approaches to the
  maintenance of a large controlled medical
  terminology. J Am Med Inform Assoc. Jan 1994-28
  Feb 1994 1(1)35-50 ISSN 1067-5027.
• Cimino J.J. From data to knowledge through
  concept-oriented terminologies experience with
  the Medical Entities Dictionary. J Am Med Inform
  Assoc. May 2000-30 Jun 2000 7(3)288-97 ISSN
  1067-5027.
• Haux R., Department of Medical Informatics,
  University of Heidelberg, Institute for Medical
  Biometry and Informatics, Germany. Health and
  medical informatics education perspectives for
  the next decade. International Journal of Medical
  Informatics. Jun 1998 50(1-3)7-19.
• McDonald C.J., Overhage J.M., Dexter P., Takesue
  B., and Suico J.G., Indiana University School of
  Medicine, Regenstrief Institute for Health Care,
  Indianapolis 46202, USA. What is done, what is
  needed and what is realistic to expect from
  medical informatics standards. International
  Journal of Medical Informatics. Feb 1998
  48(1-3)5-12.
• Reinhard O., Ohr C., Schmucker P., and Haux R. On
  stepwise integrating an electronic patient record
  based on digital-optical archiving and
  multi-purpose health professional workstations
  experiences at the Heidelberg University
  Hospital. Medinfo. 9 Pt 189-92, 1998.