Presentation to the Wonca International Classification Committee (WICC) during the WICC - CISP Club open seminar, Lyon (France), 31/08/2017

1. Open Seminar
Wonca – Université Claude Bernard Lyon 1
Lyon 31/08/2017
Risk management and
classification:
Why should the future be told in
a different language?
Philippe AMELINE
philippe@ameline.net / @p_ameline

2. The issue
1) create dedicated codes
In order to enter the risk management
universe, is it better to...
à la “risk of cardiovascular disease”
...or to...
2) use existing codes in a new dedicated
structure slot
Episode of
Risk of “cardiovascular disease”
When dealing with risk, I have witnessed that the usual behavior is to
create a set of specific codes (or terminology entries, or ontology
concepts, etc) à la “risk of cardiovascular disease”, as if risk
management was a separate “module” (in the way drug management
and lab results are so often treated as separate components with
their own internal languages, say ATC and LOINC).
The alternative is to keep using the codes usually allocated for
diseases and create a new structure element for risk like it is usually
done for describing a processes the episodes of these diseases.
We will discuss the two options from a nearly twenty years hindsight
with the Ligne de Vie project.

3. What did we learn with the
Ligne de Vie ?
Hence, when telling Mrs Smith’s “health journey”,
creating dedicated “risk of” codes obviously would
mean that all diseases codes are to be duplicated!
Tell and organize a patient health journey:
- 55000 terms ontology (the words)
- documents as trees (the (dependancy) grammar)
The Ligne de Vie has been created (mainly inside the Cisp Club) to
tell and organize the health journey of individuals. Here, an acute
pancreatitis led to a diabetes later treated by insulin.
To tell this narrative, all documents (notes, reports… including
demographics) are trees filled with elements form a 55,000 terms
ontology... in the same way an English sentence is a grammar tree
filled with words from the English vocabulary.
In 2003 we started working on risk management for cancer with a
group of family practitioners whose main asset is a “dictionary of
consultation results”.
They immediately added “risk of...” entries for the 8 cancers at stake
(mouth, cervix, breast...) in their dictionary to the existing 270
entries.
When I realized that duplicating each disease concept would lead
me to add and maintain thousands of entries in the Ligne de Vie
ontology, I decided to instead create a dedicated structure. I now
realize how deeply this initial decision made both partners of this
project radically part.

4. ...
Risk management open the way to a new paradigm
- “now” takes place on the middle of the screen
- risk targets are shared in the patient frame
The good old “patient record” becomes a
Health Project Manager
Implementing risk management in the Ligne de Vie as a “virgin land
to explore” allowed us to discover 3 main things:
1) Medical records are fully dedicated to the patient history while risk
management is entirely future oriented. The “now” cursor moved
accordingly from the right side to the middle of the screen.
2) Medical records operate in the practitioner’s reference frame while
risk must be managed in the patient’s reference frame. For
example, would the patient’s weight to be measured every day,
this target has no place to be in doctor’s workflow when this
practitioner only sees the patient once every month.
3) Medical records are local tools while risk management is by nature
project oriented since its processes must be shared by a team (at
least the patient and her GP).
You can easily guess that the conceptual distance between us and
the group of GPs who managed risk as “yet another consultation
result greatly increased with each step.

5. From SOAP to (PRA)SO
If a ? remains, the document can be shared with
other health team members as a “virtual staff”.
From local to team work
Let’s go further and imagine a patient encounter user interface that
fits with such concepts.
The “now” separator line “slides open” to display a “cognitive map”
that separate the previous processes on the left and the next ones
on the right side. Reasons for encounters and clinical observations
are recorded on the cognitive map.
Here, a set of symptoms and observations lead the practitioner to
think that a new problem is arising and/or that the treatment must be
adapted. He added two question marks on the right side.
If some ‘?’ remain unlabeled (unanswered) when the encounter ends,
then the practitioner can share this cognitive map with other
members of the patient team and use group decision tools (like the
Question, Options, Criteria (QOC) graph) to adapt the project.
One of the most precious achievement of such maps is to keep an
history of decision making (at least to answer the usual question
“why in hell did they prescribe/stop this?” in a way that can allow to
smartly compare the previous context with current situation).

6. New paradigms of care
Cohen, J (21st Century Challenges for Medical Education; 9th International Medical Workforce Conference;
Melbourne, Australia; November 2005)
The individual → The community
Acute disease dominates → More chronic illness /disability
Episodic care → Continuous care
Cure of disease → Preservation of health
Reactive → Prospective
Physician provider → Teams of providers
Paternalism → Partnership with patients
Provider centred → Patient / family centred
Parochial health threats → Global health threats
Jordan J Cohen was the president of the US Association of Medical
Colleges when he published this list of new paradigms of care.
Ten years later, we can assess that the medical domain is still stuck
in the same old paradigm, delineated on the left… and I would say
that medical information systems are not fit – by far - to make the
transition easier. In many countries the electronic health record
mainly evolves as a surveillance system, usually motivated by pay
for performance (P2P) considerations.
Risk management may be the proper incentive to have practitioners
set out.
Not embracing change, and keeping software whose decision
support role marginally evolved since 1980, could lead medical
doctors to deadly miss two major turns: the switch from acute care to
chronic follow up and the global evolution toward the information
society.

7. Conclusion
There is little reason to create dedicated codes
except if trying to insert an ersatz of risk
management inside (current / old style) information
systems.
Risk management is a real evolutionary path both
for medicine and health information systems.
Telling the future in a different language is a good
way to get stuck in the old paradigm.
Risk management has a larger scope than usually
thought, for example drug interaction is RM.
John Hagel is an author whose inspiring book “The Power of Pull”
subtitles “How small moves, smartly made, can set big things in
motion.” I believe, from what I experienced with the Ligne de Vie that
embracing risk management can be such small move to pull
practitioners out of their (deadly) comfort zone toward Cohen’s
paradigm shift.
Unfortunately, there is an other side to this coin, and small resistance
to change, for example by packaging new concepts so that they can
fit in old information systems, can quickly lead to obsolescence..
Don’t try to fold the future into the old way. Embrace change and use
risk management as an opportunity to become the managers of your
patients’ health projects… before modern information systems
enable other people to enter the cast for this role