Point in time

Hi all.

I need some help in understanding the concept 'point in time'. As far
as I understand it now the completion date/time of an event (f.i. a
blood pressure measurement) is registered as a point in time in the
event section of the archetype.

Thing is that I expected to find an entry field where I can enter
that particular date and time. In the EHR_IM I found the Action
class (p.71) which purpose seems to be to capture/record those points
in time, is I figured that's the place to be.

So I went back and looked into the template blood pressure developed
by the NHS_UK (.../knowledge/templates/dev-uk-nhs/html/Blood%
20pressure.html).

There one finds the annotation 'event' (any event) nicely
illustrated with a clock. But still no 'place' where one could enter
the date and time of the event that took place.

Can somebody please explain to me how this works?

Cheers,

Stef Verlinden

Stef,
In the LOINC timing axis point in time means that the measurement was
done at a specific time point. I'd be interested to know how this is
different. Ssusan

From the LOINC Manual:

"One can either measure a property at a moment (point) in time or
measure it over a time interval and integrate, in the mathematical
sense, over time. In the latter case, we aggregate a "series" of
physiologic states into a single scalar value that reflects some
"average" property measured over the specified time interval. Intervals
also have relevance for rate measurements such as excretion (substance
rate or mass rate) or clearances (volume rates). The amount over an
interval is often expressed as a mass rate (MRAT, e.g. g/24h) or a
substance rate (SRAT, e.g. mol/24h). Interval measurements often apply
to urine and stool (e.g. collection over 24 hours and calculation of a
concentration, total amount, or clearance). They also apply to clinical
measurements such as urine outputs where we have shift totals and
24-hour totals. Event counts on physiologic monitors, such as the
number of premature ventricular contractions (PVCs) over 24 hours on a
Holter monitor, are also of this type."

Susan Matney, R.N., M.S.
email: susan.matney@siemens.com
remote office: (801) 298-4642
voice mail: (610) 219-1941
cell: (801) 680-2161

I need some help in understanding the concept 'point in time'. As far
as I understand it now the completion date/time of an event (f.i. a
blood pressure measurement) is registered as a point in time in the
event section of the archetype.

yes, that's my understanding as well.

Thing is that I expected to find an entry field where I can enter
that particular date and time. In the EHR_IM I found the Action
class (p.71) which purpose seems to be to capture/record those points
in time, is I figured that's the place to be.

well, there is a 'time' field in Action which is where you would record
the time of an action. But blood pressure is an observation, and
unlike Action, all observations are modeled as a HISTORY of events.

So in OBSERVATION, we find a HISTORY that has an 'origin'
time and also contains a set of time EVENT's.

Each EVENT also has a 'time' field (with some restrictions on how
it relates to 'origin' in the containing HISTORY).

However, you will rarely find any reference to one of these fields in
an ADL document because these fields are generally unconstrained (i.e you
wouldn't write an archetype that said the blood pressure had to be
measured at 03-06-2007). You will find that the 'offset' field is often used -
this is defined as the difference in time between the EVENT and the
base time in the HISTORY. It can be used to model EVENT's that
must occur at set times after the 'origin' of the HISTORY (apgar etc).

INTERVAL_EVENT can be used to record fancy averages over time
etc.

(I don't write archetypes so take this with a grain of salt! It's my
understanding
from a technical background, not a clinical background)

Andrew

Thanks for your clear explanation but it doesn't provide the clue I'm
looking for. Since I lack a most of the technical background I can
only approach it from a clinical point of view.

My question is: Where/how do I register the data and time of an event
that took place.
So let's say that at this moment (May 18 2007 @10hrs10min) I want to
register an event that took place last night (May 17 2007 @23hrs53min).
How/where in the AT can I create a data entry field for this (if
that's necessary at all) and how would this look like in a data entry
template html page

Stef Verlinden wrote:

Thanks for your clear explanation but it doesn't provide the clue I'm
looking for. Since I lack a most of the technical background I can
only approach it from a clinical point of view.
  

Andrew Patterson's understanding is correct - usually you only archetype
the offset, since you never know in advance the actual start time of the
History - e.g. when the first BP sample is made, or when birth occurs
(the origin point for Apgar scores) etc

My question is: Where/how do I register the data and time of an event
that took place.
So let's say that at this moment (May 18 2007 @10hrs10min) I want to
register an event that took place last night (May 17 2007 @23hrs53min).
How/where in the AT can I create a data entry field for this (if
that's necessary at all) and how would this look like in a data entry
template html page
  

Stef, you don't do this in archetypes or templates, since both of these
artifacts are used as a kind of standardised model of data (and data
capture), not a data instance. In the run time system, when archetypes
and templates are used at point of data entry, the times will be (should
be) filled in by the software....

hope this helps.

- thomas

Stef Verlinden wrote:

Thanks for your clear explanation but it doesn't provide the clue I'm
looking for. Since I lack a most of the technical background I can
only approach it from a clinical point of view.

Andrew Patterson's understanding is correct - usually you only
archetype
the offset, since you never know in advance the actual start time
of the
History - e.g. when the first BP sample is made

Not in advance but the moment you 'make' the first BP sample you
(could) know exactly when that occurred.

, or when birth occurs
(the origin point for Apgar scores) etc

My question is: Where/how do I register the data and time of an event
that took place.
So let's say that at this moment (May 18 2007 @10hrs10min) I want to
register an event that took place last night (May 17 2007
@23hrs53min).
How/where in the AT can I create a data entry field for this (if
that's necessary at all) and how would this look like in a data entry
template html page

Stef, you don't do this in archetypes or templates, since both of
these
artifacts are used as a kind of standardised model of data (and data
capture), not a data instance. In the run time system, when archetypes
and templates are used at point of data entry, the times will be
(should
be) filled in by the software....

This part I still don't understand. So you're saying that you don't
record the actual date/time of measurement but only the data entry
date/time, the latter being provided by the software.
What if an event is entered at a later point of time and it's
necessary to know the actual time of measurement?

Stef

Just to elucidate:

In relation to observations there are 3 times that are formally captured:

  1. The timing event - this is the time the sample is taken with blood tests, or the time of measurement of a blood pressure. This is captured as part of the history class.
  2. The event context of the composition - this relates to the clinical time, report time or whatever.
  3. The commit time - when the record is committed to the EHR.

The important thing is that you could commit a record on Monday having seen someone on Saturday (and made handwritten notes) and recorded a blood sugar they took on Friday at 10.30am.

Hope this helps…Sam

Stef Verlinden wrote:

In the run time system, when archetypes
and templates are used at point of data entry, the times will be
(should
be) filled in by the software....

Let me give an example why we would like to have a separate data
measurement and data entry date and time field.
For instance in the 'youth' healthcare many health care providers are
still working with paper records to register growth and weight
progression (and these data point are then plotted by hand in a
graph). If we would like to 'persuade' these healthcare providers to
start using an electronic system (based on openEHR), they would want
to be able to 'import' their existing data. If now only the data
entry date/time but not the data measurement date/time are
registered, there's not much sense in doing so, since no age/weight
and/or age/length plots can be generated anymore. These plots are
valuable tools when monitoring the progression of children.

How would one be able to do such a thing using the current AT structure?

Regards,

Stef

Absolutely. This was what I was looking for.

Thanks,

Stef

Absolutely - there is no restriction on the date of the measurement of data - we do this sort of thing all the time…Sam

Stef Verlinden wrote: