Referent Tracking: The New Paradigm Dagstuhl May 23th, 2006

1
Referent Tracking The New ParadigmDagstuhl
May 23th, 2006

• Werner Ceusters, MD
• Ontology Research Group
• Center of Excellence in Bioinformatics Life
  Sciences
• SUNY at Buffalo, NY

2
Three models for a Comprehensive Patient Care
Information System
?
Information model of the medical record What can
be said ?
inferences
requirements
Process model of clinical care What ought to
occur ?
Inferred model of the state of the patient What
actually occurs.
Inferred model of the state of the patient What
actually occurs.
modification
Rector AL, Nolan WA, and Kay S. Foundations for
an Electronic Medical Record. Methods of
Information in Medicine 30 179-86, 1991. (Figure
1)
3
An event model of medicalinformation
representation

• The real world consist of objects (or entities)
  
• Objects interact with other objects and can be
  associated with other objects by relationships
• When two or more objects interact in the real
  world, an event is said to have occurred

What event occurred in my left hand being a
part of my left arm ?
Huff SM, Rocha RA, Bray BE, Warner HR, and Haug
PJ. An Event Model of Medical Information
Representation. J Am Med Informatics Assoc.
19952116-134.
4
Observation Event Instance Created from the
Statement Surgical clips are again seen along
the right mediastinum
?
Huff SM, Rocha RA, Bray BE, Warner HR, and Haug
PJ. An Event Model of Medical Information
Representation. J Am Med Informatics Assoc.
19952116-134.
5
What is wrong in both cases(and so many others) ?

• Further clues to the answer
• Huff SM, Rocha RA, Bray BE, Warner HR, and Haug
  PJ. An Event Model of Medical Information
  Representation. J Am Med Informatics Assoc.
  19952116-134.
• Rector et al.s paper Many of the difficulties
  experienced in attempting to generalize existing
  patient record systems stem from the fact that
  they have pre-selected and distorted information
  in order to fit into particular applications,
  usually clinical research and epidemiology. The
  models emit much of the information actually used
  in clinical care and do not accurately reflect
  the real status of the data they record.

6
What is wrong (continued) ?

• An overemphasis on data and information and too
  little attention to reality
• data modelling
• information modelling
• Is the Object Oriented model approach any
  better, since, after all, objects are said to be
  those things that exist in reality ?
• The object-oriented model is based on a
  collection of objects
• An object contains values stored in instance
  variables within the object.
• Unlike the record-oriented models, these values
  are themselves objects.

7
Current mainstream thinking
8
The story of Jane Smithan old case, well known
in the literature ...
9
July 4th, 1990 Jane goes shopping
10
A visit to the hospital

• City Health Centre Dr. Peters
• (City HC) Dr. Longley

11
Diagnosis a severe spiral fracture of the femur
12
The City HCs medical record

• Main principles
• a faithful record of the clinicians
  observations what they have heard, seen, thought
  and done.
• captures in a structured form all of the
  clinically significant information in the
  narrative notes, where by clinically significant
  they mean the information which is within the
  medical domain rather than the domain of everyday
  life.

Rector AL, Nowlan WA, Kay S, Goble CA, Howkins
TJ. A framework for modelling the electronic
medical record. Methods Inf Med. 1993
Apr32(2)109-19.
13
CityHCs representation formalism(for statements
in records)
Categories represent concepts and are analogous
to classes in other formalisms
Individuals concrete instances of categories
which persist in space and time
Occurrences are specific occurrences of
individuals and must be situated in space and
time. The most important group of occurrences are
observations i.e. agents observations of
individuals.
Rector AL, Nowlan WA, Kay S, Goble CA, Howkins
TJ. A framework for modelling the electronic
medical record. Methods Inf Med. 1993
Apr32(2)109-19.
14
A look at the database use of SNOMED codes for
unambiguous understanding
How many numerically different disorders are
listed here ?

How many different types of disorders are listed
here ?

How many disorders have patients 5572, 2309 and
298 each had thus far in their lifetime ?

cause, not disorder
15
Does seeing the labels help you ?
16
Main problem areas for CityHCs EHR

• Statements refer only very implicitly to the
  concrete entities about which they give
  information.
• Idiosyncracies of concept-based terminologies
• tell us only that some instance of the class the
  codes refer to, is refered to in the statement,
  but not what instance precisely.
• Are usually confused about classes and
  individuals.
• Country and Belgium.
• Mixing up the act of observation and the thing
  observed.
• Mixing up statements and the entities these
  statements refer to.

17
Consequences

• Very difficult to
• Count the number of (numerically) different
  diseases
• Bad statistics on incidence, prevalence, ...
• Bad basis for health cost containment
• Relate (numerically same or different) causal
  factors to disorders
• Dangerous public places (specific work floors,
  swimming pools),
• dogs with rabies,
• HIV contaminated blood from donors,
• food from unhygienic source, ...
• Hampers prevention
• ...

18
Proposed solutionReferent Tracking

• Purpose
• explicit reference to the concrete individual
  entities relevant to the accurate description of
  each patients condition, therapies, outcomes,
  ...
• Method
• Introduce an Instance Unique Identifier (IUI) for
  each relevant individual ( particular,
  instance).
• Distinguish between
• IUI assignment for instances that do exist
• IUI reservation for entities expected to come
  into existence in the future

19
Ontology

• Ontology the study of being as a science
• An ontology is a representation of some
  pre-existing domain of reality which
• (1) reflects the properties of the objects within
  its domain in such a way that there obtains
  a systematic correlation between reality and the
  representation itself,
• (2) is intelligible to a domain expert
• (3) is formalized in a way that allows it to
  support automatic information processing
• ontological (as adjective)
• Within an ontology.
• Derived by applying the methodology of ontology
• ...

20
An ontological analysis
continuants
21
Ontological recategorisation
Jane Smiths consultation with Dr. Peters
at City HC on 4th July 1990
Dr. Peters assessment of Jane Smiths fracture
of femur at City HC on 4th July 1990
22
Essentials of Referent Tracking

• Generation of universally unique identifiers
• deciding what particulars should receive a IUI
• finding out whether or not a particular has
  already been assigned a IUI (each particular
  should receive maximally one IUI)
• using IUIs in the EHR, i.e. issues concerning the
  syntax and semantics of statements containing
  IUIs
• determining the truth values of statements in
  which IUIs are used
• correcting errors in the assignment of IUIs.

23
Architecture of aReferent Tracking System (RTS)

• RTS system in which all statements referring to
  particulars contain the IUIs for those
  particulars judged to be relevant.
• Ideally set up as broad as possible
• some metrics
• of particulars referred to by means of IUI
• of HCs active in a region
• Geographic region
• functional region defined by contacts amongst
  patients
• of patients referred to within a region
• Services
• IUI generator
• IUI repository statements about assignments and
  reservations
• Referent Tracking Database (RTDB) index (LSID)
  to statements relating instances to instances and
  classes

24
IUI generation

• Universally Unique IDs
• recently standardized through ISO/IEC
  9834-82004,
• specifies format and generation rules enabling
  users to produce 128-bit identifiers that are
  either guaranteed or have a high probability of
  being globally unique
• Meaningless strings
• Central management or certification not needed to
  guarantee uniqueness
• (But use as IUI requires this)

25
IUI assignment

• an act carried out by the first cognitive
  agent feeling the need to acknowledge the
  existence of a particular it has information
  about by labelling it with a UUID.
• cognitive agent
• A person
• An organisation
• A device or software agent, e.g.
• Bank note printer,
• Image analysis software.

26
Criteria for IUI assignment (1)

• The particulars existence must be determined
• Easy for persons in front of you, body parts, ...
• Easy for planned acts they do not exist before
  the plan is executed !
• Only the plan exists and possibly the statements
  made about the future execution of the plan
• More difficult subjective symptoms
• But the statements the patient makes about them
  do exist !
• However
• no need to know what the particular exactly is,
  i.e. which universal it instantiates
• No need to be able to point to it precisely
• One bee out of a particular swarm that stung the
  patient, one pain out of a series of pain attacks
  that made the patient worried
• But this is not a matter of choice, not any
  out of ...

27
Criteria for IUI assignment (2)

• The particulars existence may not already have
  been determined as the existence of something
  else
• Morning star and evening star
• Himalaya
• Multiple sclerosis
• May not have already been assigned a IUI.
• It must be relevant to do so
• Personal decision, (scientific) community
  guideline, ...
• Possibilities offered by the EHR system
• If a IUI has been assigned by somebody, everybody
  else making statements about the particular
  should use it

28
Assertion of assignments

• IUI assignment is an act of which the execution
  has to be asserted in the IUI-repository
• Di ltIUId, Ai, tdgt
• IUId IUI of the registering agent
• Ai the assertion of the assignment lt IUIp, IUIa,
  tapgt
• IUIa IUI of the author of the assertion
• IUIp IUI of the particular
• tap time of the assignment
• td time of registering Ai in the IUI-repository
• Neither td or tap give any information about when
  IUIp started to exist ! That might be asserted
  in statements providing information about IUIp
  .

29
Management of the IUI-repository

• Adequate safety and security provisions
• Access authorisation, control, read/write, ...
• Pseudonymisation
• Deletionless but facilities for correcting
  mistakes.
• Registration of assertion ASAP after IUI
  assignment
• (virtual, e.g. LSID) central management with
  adequate search facilities.

30
Representation in the EHR

• Relevant particulars referred to using IUIs
• Relationships that obtain between particulars at
  time t expressed using relations from an ontology
  (type OBO)
• Statements describing for each particular, at
  time t
• Of what universal from an ontology it is an
  instance of
• AND/OR (if one insists)
• By means of what concept from a concept-based
  system it can sensibly be described

particulars
31
PtoP statements - particular to particular

• ordered sextuples of the form Ri ltIUIa,
  ta, r, o, P, trgt

IUIa is the IUI of the author of the statement,
ta a reference to the time when the statement
is made, r a reference to a relationship
(available in o) obtaining between the
particulars referred to in P, o a reference to
the ontology from which r is taken, P an ordered
list of IUIs referring to the particulars between
which r obtains, and, tr a reference to the
time at which the relationship obtains.

• P contains as much IUIs as required by the
  arity of r. In most cases, P will be an ordered
  pair such that r obtains between the particular
  represented by the first IUI and the one referred
  to by the second IUI.
• As with A statements, these statements must
  also be accompanied by a meta-statement capturing
  when the sextuple became available to the
  referent tracking system.

32
PtoU statements particular to universal
Ui ltIUIa, ta, inst, o, IUIp, u, trgt IUIa is
the IUI of the author of the statement, ta a
reference to the time when the statement is made,
inst a reference to an instance relationship
available in o obtaining between p and cl, o
a reference to the ontology from which inst and
u are taken, IUIp the IUI referring to the
particular whose inst relationship with u is
asserted, u the universal in o to which p
enjoys the inst relationship, and, tr a
reference to the time at which the relationship
obtains.
33
U--tuples negative findings

Relation
Type of Negative F
inding

Examples



type

he denies abdominal pain no alcohol abuse
C1

ltp, ugt

A particular is not related in a
85.4

specific way to any instance of a
no hepatosplenomegaly he has no children,
without any cyanosis

universal at some given time

which ruled out primary hyperaldosteronism,
C2

ltp, ugt

A par
ticular is not the instance of
12.4

a given class at some given time

nontender, in no apparent distress, Romberg
sign was absent
,
no palpable lymph nodes

this record is not available to me it is not
C3

ltp, pgt

A particular is not related to
2.2

another partic
ular in a specific
the intense edema she had before he has not
way at some given time

identified any association with meals.


p ranges over particulars, u over universals

34
PtoCO statements particular to concept code
Coi ltIUIa, ta, cbs, IUIp, co, trgt IUIa is
the IUI of the author of the statement, ta a
reference to the time when the statement is made,
cbs a reference to the concept-based system
from which co is taken, IUIp the IUI referring
to the particular which the author associates
with co, co the concept-code in cbs which the
author associates with p, and, tr a reference
to the time at which the author considers the
association appropriate,
35
Interpretation ofPtoCO statements

• must be interpreted as simple indexes to terms in
  a dictionary.
• All that such a statement tells us, is that
  within the linguistic and scientific community in
  which cbs is used, the terms associated with co
  may - i.e. are acceptable to - be used to denote
  p in their determinative version.

36
A SNOMED-CT example

• ltIUI-0945, 18/04/2005, SNOMED-CT v0301,
  IUI-1921, 367720001, forevergt
• IUI-0945 author of the statement
• IUI-1921 the left testicle of patient
  IUI-78127
• 367720001 the SNOMED concept-code to which left
  testis is (in SNOMED) attached as term
• So we can denote IUI-1921 by means of
• that left testis
• that entire left testis
• that testicle, that male gonad, that testis
• that genital structure
• that physical anatomical entity
• BUT NOT that SNOMED-CT concept

37
Pragmatics of IUIs in EHRs

• IUI assignment requires an additional effort
• In principle no difference qua (or just a little
  bit more) effort compared to using directly codes
  from concept-based systems
• A search for concept-codes is replaced by a
  search for the appropriate IUI using exactly the
  same mechanisms
• Browsing
• Code-finder software
• Auto-coding software (CLEF NLP software Andrea
  Setzer)
• With that IUI comes a wealth of already
  registered information
• If for the same patient different IUIs apply, the
  user must make the decision which one is the one
  under scrutiny, or whether it is again a new
  instance
• A transfert or reference mechanism makes the
  statements visible through the RTDB

38
PtoN-statements

• Nilt IUIa, ta, ntj, ni, IUIp, trgt
• The person referred to by IUIa asserts at time ta
  that ni is the name of the nametype ntj that
  designates in the real world the particular
  referred to by IUIp at tr. This template will
  further be referred to as PtoN template.
• Would assert that Werner is my first name, and
  Ceusters is my last name.

39
Advantage better reality representation
IUI-003
40
Other Advantages

• mapping as by-product of tracking
• Descriptions about the same particular using
  different ontologies/concept-based systems
• Quality control of ontologies and concept-based
  systems
• Systematic inconsistent descriptions in or
  cross terminologies may indicate poor definition
  of the respective terms

41
A case study

• Double goal
• Application of referent tracking to a concrete
  patient story
• Ontological analysis of what is involved
• The latter is NOT to be performed to the same
  extent when referent tracking is used as an
  alternative to coding using concept-based
  systems.
• So, dont go home with the idea thats all too
  cumbersome and time cosuming

42
Jim Ciminos Woods Hole case

• Jane Smith is a 30 year old, Native American
  female who presents to the emergency room with
  the chief complaint of cough and chest pain.
• The patient reports that she has had a productive
  cough for three days but that chest pain
  developed one hour ago.  She gives a history of
  hypertension.  She also reports that she was
  treated in the past for tuberculosis while she
  was pregnant.  The patient reports an allergy to
  Bufferin.
• Physical examination revealed a well-developed,

43
Step 1 identify the phrases referring to
particulars

• Jane Smith is a 50 year
  old ,
• Native American female who presents
• to the emergency room
• with the chief complaint
• of cough and chest pain.

44
Step 2 indentify to what particulars these
phrases refer
Jane Smith
Jane Smiths age
Jane Smiths race
Jane Smith
Jane Smiths gender
Jane Smiths showing up at ...
A specific emergency room of health facility XYZ
A specific pain experienced by Jane Smith
Jane Smiths complaining primarily about ...
A temporal part of Jane Smiths life marked by
happenings of coughs
Jane Smiths chest
45
Compare with simple clinical coding in
juxtaposition
46
Compare with the output of the perfect semantic
analyser we all would dream of
Compare with the output of the NAIVE !!! semantic
analyser we all would dream of
CS3-complaining
47
What it (more or less) should be
chest-pain
CS3-complaining
Has-Saying
Has-referent
CS3-chest pain
Has-Saying
coughing
Has-referent
CS3-coughing
48
Conclusion

• Referent tracking can solve a number of problems
  in an elegant way.
• Existing (or emerging) technologies can be used
  for the implementation.
• Old technologies (concept-based systems) can play
  an interesting, but different role.
• Big Brother feeling is to be expected but with
  adequate measures easy to fight.
• The proof of the pudding is in the eating
• Pilote is going to be set up