Confusing naming of OBSERVATION and EVALUATION?

Clinical (archive)
1

Hi!

I just want to get some more global response to a little discussion
some of us had here at LiU. So here comes some provocations... :slight_smile:

In contact with clinicians we have discovered that the naming of the
openEHR reference model classes OBSERVATION and EVALUATION (and of
course the corresponding archetype families) is confusing partly
because of the class names themselves. The distinction in practical
clinical documentation cases is often far from obvious when just
looking at the names, ontology or the "investigator-patient-cycle"
(see e.g. http://www.openehr.org/wiki/display/ontol/Ontologies+Home )
.

When one starts telling clinicians _not_ to focus on the names and
explains the _technical_ difference of available attributes and
mechanisms of the respective classes, then the confusion often gets
reduced.

What OBSERVATION does provide is extensive support for structured
reporting of timed events. EVALUATION on the other hand is the most
generic concrete form of CARE_ENTRY.

The problem is that the words "observation" and "evaluation" already
are so loaded with meaning. Maybe some renaming (at least during
pedagogical presentations) would help understanding. Mentally rename:
CARE_ENTRY to ABSTRACT_CARE_ENTRY (usually never seen by clinicians)
EVALUATION to CARE_ENTRY
OBSERVATION to CARE_ENTRY_SUPPORTING_STRUCTURED_TIMING

(I'm not insisting on actual technical changes to class names, but
rather in mental presentation approach. I think the openEHR RM-design
is great, it's the presentation and naming I'm concerned about)

This way of thinking may make it easier to select class according to
documentation (and reuse) needs rather than trying to fit it into some
more or less useful ontological thinking.

If you want to create EHR systems with advanced graphical user
interfaces or decision support using timepoints in the data, then the
task is a lot easier if there is a uniform way of reporting time (as
in the OBSERVATION class) than if people squeeze in timepoints in
arbitrary places of EVALUATION, so there is a technical need for both
classes. The selection of class during archetype design could however
rather be based on the possibility/applicability of timing-structure
rather than ontological hair-splitting (Don't get me wrong, I do
appreciate ontologies in a general sense...).

INSTRUCTION and ACTION usually lead to less confusion, and are
probably less harmful names.

The focus on the investigator-patient cycle and the "beauty" of an
"ontological" approach is probably attractive at first, but if it
leads to problems during presentation and modelling, then their role
should be played down.

Best regards,
Erik Sundvall
erisu@imt.liu.se http://www.imt.liu.se/~erisu/ Tel: +46-13-227579

P.s. Compare to the deceptive "beauty" and simplicity of the handful
of HL7 v3 RIM core classes and then consider if the
openEHR-cycle-and-ontology-presentation-focus is a similar mental trap
that seems nice at first, but really does not make modelling and reuse
simpler. (...or was that too provocative...)

2

Erik,

I understand the problem of ‘when to use Observation and when to use Evaluation’.
Like all of us it is sometimes confusing.

But not any longer for me at least.

Observation is a statement about a status in the patient system.
That what can be heard, seen, smelled, tasted, touched by our senses and measured by a device.

Evaluation is a statement with an inference about a process in the patient system.
That what we infer using Observations, our knowledge and experience: e.g. diagnosis, problem, risk, goal, …

Patient system:
-Body: relating to anatomy, organ system, physiology, genetics, chemistry
-Person as a whole. e.g. wight, height, mental status
-Social system: family, neighbors, work
-Extended (social) system: parts of society, environment, etc

Processes lead to states that can be observed.
States are consequences of processes.
Only via observations of states we can make inferences about the processes using our knowledge and experiences.

In the same fashion Instructions and Actions can be defined.

Gerard

– –
Gerard Freriks, MD
Huigsloterdijk 378
2158 LR Buitenkaag
The Netherlands

T: +31 252544896
M: +31 620347088
E: gfrer@luna.nl

Those who would give up essential Liberty, to purchase a little temporary
Safety, deserve neither Liberty nor Safety. Benjamin Franklin 11 Nov 1755

3

Hi Erik

This is very useful and something we have come to do as well. I believe that
the idea of state at the time of measurement is also important in the
OBSERVATION class, not just the timing itself.

Cheers, Sam

4

Building on what has gone before....

In speaking with clinicians I seem to have found that we need to explain
it as a mix of clinical use and the RM attributes.

Instructions and actions are reasonably simple to isolate.
The confusion is between OBSERVATION and EVALUATION almost universally.

So, assume that the default archetype should be considered to be an
EVALUATION unless the modeller can answer yes to one or more of the
following:

Is this recording the history (something that the patient has told you
has happened) or examination for a patient (touch, feel, smell etc)?
Is this recording a measurement? eg BP, weight, lab result
Is this recording something that requires a context/state to interpret
the finding appropriately? eg fasting state for glucose reading
Is this recording something that requires data to be recorded against a
specific event eg weight at birth?
Is this recording a change in parameters eg loss, gain over time?
Is this recording something over an interval?

If the answer to one or more of these questions is yes, then the data
should be modelled as an OBSERVATION. Most often the modeller finds
that the data fits more than one of these criteria which confirms that
OBSERVATION is correct choice.

My 5c

Heather

Sam Heard wrote:

5

Dear folks,

rarely I think I can do any contribution to the list, especially because my native language isn´t English, but as as psychiatrist the distinction between the two concepts is very clear..

The psychiatric assessment is divided in two parts, one is the Observation= semiotics (signs and symptoms, test results, etc|= phenomenological OBSERVATION) EVALUATION for us is the interpretation of all the things we observed, we classify that in a syndrome, which leads our diagnosis and treatment plan.

Following some info about phenomenology… FYI

Edmund Husserl is considered the founder of phenomenology and can help us to understand the boundaries between the tow concepts.

" Husserl aimed to develop an objective foundation for all of human knowledge. As he wrote in Crisis of the European Sciences (1936), he was frustrated with the prejudiced way philosophers and psychologists conduct their research. They start out, he complained, already ‘knowing’ what kinds of observations count as evidence for their conclusions. Yet they have no clear understanding of the relationship between observation, evidence, and conclusions.

Husserl viewed his phenomenological method as an improvement on the work of René Descartes in his Meditations on First Philosophy. Descartes wanted to find a principle that could guarantee the certainty of human knowledge. He proceeded by purposely doubting all of his faculties, hoping he would find something he could not doubt. ‘I sense the world,’ he thought to himself, ‘but I cannot conclude that my perceptions are true. I apply mathematical formulas, but I cannot conclude that they describe the real world. I hold religious beliefs, but I cannot assume that they describe ultimate reality.’ Descartes found he could not doubt the ‘I’ making these assertions. From this one certainty, Descartes deduced many others.

Husserl wrote in his own book Cartesian Meditations (1931) that he was extremely impressed with Descartes’ method of doubt. However, he thought Descartes did not spend enough time using it. Instead of learning how his consciousness works, Descartes jumped too quickly to find certainty. In the end, Descartes only reaffirmed all of the old prejudices about the organisation of knowledge. Husserl hoped to do better.

Husserl developed a toolkit of concepts to make the practice of doubt productive. Most important is the practice he named ‘phenomenological reduction’. ‘Phenomena’, as defined by Immanuel Kant, are appearances in consciousness, that is, perceptions, thoughts, and images, rather than things in themselves. When Husserl practiced the phenomenological reduction, he reduced everything in consciousness to the status of appearances. He contrasted the phenomenological reduction with the ‘natural attitude’, that is, our everyday mode of thinking. In the natural attitude we view our thoughts as windows onto the world. For example, we might say to ourselves, ‘I see a tree limb in the road; I’d better not hit it. I feel sick; is there something wrong with my body?’ In the phenomenological reduction, we view the same thoughts as events in consciousness: ‘I see a tree limb; I plan to avoid it; I feel sick; I hope to be well.’ Using phenomenological observation followed by reflection, Husserl developed additional concepts with which to organise and describe the events of consciousness.

Husserl believed that the observation of pure consciousness could clarify the relationships between philosophical theories and the evidence for them. Thus, it would yield new answers to important philosophical questions. In Ideas (1913) he presents some of his answers. For example, he considers the great metaphysical question of the distinction between the mental and the material. What makes them so different that philosophers have often found them to be two irreducible categories? Under phenomenological reduction they appear in consciousness as two distinct types of objects. Material objects appear as a series of perspectives, that is, different views from different sides, all of which are referred to a single object of consciousness. Each mental object, however, appears to be given whole all at once, and related in time to other mental objects.

Husserl never completed his work of systematically articulating the relationships between observation, evidence, and conclusions (EVALUATION). Yet every step along the way, he discovered features of consciousness that he found enormously exciting. His passion and intellectual depth can sometimes be hard to see, however, because his writing style is downright turgid. Sadly, most of the popular criticisms of his work come from readers daunted by the technical vocabulary. Instead of engaging with Husserl’s work, they argue that the phenomenological reduction is impossible or that Husserl failed because he never reached his goal. But the viability and success of Husserl’s method are attested to by his influence. The leading continental philosophers of the twentieth century, Martin Heidegger and Emmanuel Levinas, were avid students of Husserl, applying his conceptual toolkit in original ways. Humanistic psychologists credited Husserl with teaching them to emphasise consciousness, and critical sociologists credited him with helping them recognise political distortions of knowledge.

Suggested reading
Husserl, E. 1977 [1931]. Cartesian Meditations. Dordrecht: Kluwer.
Husserl, E. 1990 [1913]. Ideas Pertaining to a Pure Phenomenology and to a Phenomenological Philosophy. Dordrecht: Kluwer.
Kohak , E. 1978. Idea and Experience: Edmund Husserl’s Project of Phenomenology in ‘Ideas I’. Chicago: University of Chicago Press