modeling non-medical data

Technical (archive)
1

I have come across an interesting opportunity to do some openehr
modeling in a sports science context. However, whilst half of
the data is medical (heart rate etc), the other half is
raw physical data (GPS location, cadence etc) related to in
this case a bike..

So I would have one large history consisting of heart rate over
time which can be modeled with existing archetypes. For the
other data (the corresponding cadence over time),
I will obviously need to construct my own
archetypes. Does anyone have any experience at modeling
this sort of non-clinical data? What would I name the
archetypes - are they in the EHR namespace? Are there any
composition archetypes suited to this non-healthcare
related data input? How does one decide what goes in
an archetype for data that comes from a bike (an archetype
for each data item, or one archetype to group the data
items together?)

(I realise it may just be easier to store that data in a
non-openehr system but doing it the openehr way has certain
attractions - some of which are that provides a unified mechanism
for all the data, and can be extended to more clinical
sports science data if that becomes important)

Andrew

2

Hi Andrew

If the data is about a person i.e. their location on a bike etc, there is no reason why it can’t go in the ehr space. There are a lot of potential entries in the EHR that are not directly clinical information but may be important in the overall context of the person’s health or what you are trying to record about it. This is one of the reasons we have the admin entry to cope with things that are not directly clinical.

regards Hugh

Andrew Patterson wrote:

3

Hi Andrew

I was not sure what cadence is…GPS is clearly location and could be plotted in an observation without any difficulty (might use the protocol to say how it was measured).

The important thing here is the state model of the medical stuff - do you want to know the exertion level etc at the time of the heart rate - or is it a parallel measurement that will provide core data for comparison?

Cheers, Sam

Andrew Patterson wrote:

4

I was not sure what cadence is...

Cadence is the RPM for the bike wheels. Other measures of
these devices are pedal torque, speed and power (some primary
measures, some computed)

http://www.saris.com/c-11-power-meters.aspx

GPS is clearly location and could be
plotted in an observation without any difficulty (might use the protocol to
say how it was measured).

I was thinking of a location observation archetype with GPS coord and
altitude, and a power observation archetype with torque, power, cadence
etc. I wasn't sure whether they should be grouped like that, or
completely separate (power archetype, torque archetype, cadence
archetype etc).

The important thing here is the state model of the medical stuff - do you
want to know the exertion level etc at the time of the heart rate - or is it
a parallel measurement that will provide core data for comparison?

The results are simultaneously measured by two devices (heart rate
through a chest monitor, bike power through a wheel monitor) and
recorded on a single master device on the bike, which can later be connected via
usb for download. However, some bike setups would measure all this
data as two separate independent streams. One could also theoretically
add in GPS tracking etc, which is currently not done by the power
measuring device (but could be in the future).

The coaches want all this data to be combined in various ways
(for instance, multiplying heart rate by average power over 15 second
windows etc). They would want it to work irrespective of how the
data got into the system.. so I want to write some archetypes that are
flexible enough to handle the various different measurement devices
and combinations of such devices.

Andrew

5

Andrew Patterson wrote:

I was not sure what cadence is...

Cadence is the RPM for the bike wheels. Other measures of
these devices are pedal torque, speed and power (some primary
measures, some computed)

No, cadence in cycling is the rotational speed (RPM) of the
cranks/chainwheel, not the bicycle wheels i.e. it is how fast you are
pedalling. How fast the wheels are turning depends what gear you are in.
Of course, on a fixed-gear, single-speed bike (eg a track bike), there
is a direct and fixed relationship between cadence and wheel RPM.

Tim C

6

No, cadence in cycling is the rotational speed (RPM) of the
cranks/chainwheel, not the bicycle wheels i.e. it is how fast you are
pedalling. How fast the wheels are turning depends what gear you are in.
Of course, on a fixed-gear, single-speed bike (eg a track bike), there
is a direct and fixed relationship between cadence and wheel RPM.

ok :-). I'm sure my cycling friends will yell at me for getting this wrong
as well - in my defence, I haven't been on a bike for more than 15 years
so my interpretation of the terminology is obviously suspect..

I'm sure I will learn a lot more by watching 'le tour' this year.

Andrew

7

I would suggest one archetype for the cycling activity - in future this might be best done with specialisations - but for the moment it would be best to put in the data as it is sampled. If the sampling frame is different then the same archetype can still be used.
Cheers, Sam

Andrew Patterson wrote:

8

But the purpose then would not be to collect data for healthcare
purposes - i.e. this post is about the use of open EHR standards and
technologies for to answer a scientific question?
Arild Faxvaag

9

But the purpose then would not be to collect data for healthcare
purposes - i.e. this post is about the use of open EHR standards and
technologies for to answer a scientific question?

Well, this is dilemna I guess - it's data about a person, created
from a physical activity - but it's not a clinical observation. It
is being used for healthcare purposes, albeit with very
specialised goals - the data is used to set safe training exertion
ranges etc for the atheletes.

I may end up doing it with an ad-hoc system - I just thought it
would be interesting to consider how it might be done using
openehr.

Andrew

10

Dear Andrew,

Our team at CHIME are part of a UK Science Council funded project to develop innovative ways of assisting coaches with athlete training.

http://www.sesame.ucl.ac.uk/

A part of our methodology is to build a centralised repository with an athlete training record and cumulative monitoring data. Some aspects of the overall data architecture are now becoming clear, others are still to be investigated (e.g. to optimise retrieval/analysis performance for high-volume data). The core athlete training record will utilise a simplified 13606/openEHR server architecture and a corresponding simplified archetype library, accessed via a web portal.

Because some of our early domain-specific materials have come from coaches who, naturally, regard their expertise as somewhat personal and private, we do not yet have permission to share the domain modelling (archetyping) work that we are doing, but I am hopeful that as the months progress we will find sufficient generic models that are not coach-specific that we can publish and share these.

With best wishes,

Dipak

11

I think it is wrong to be too 'pure' about what constitutes health
information. In this situation a heart rate archetype can be specialized
for this very specific purpose, and the cadence etc is important part of the
context of the observation, being part of a very detailed record relating to
exertion, and should be recorded in their health record as state data. I
don't see any particular problem with GPS being recorded here as well if it
fulfils a specific need.

This particular specialization may not end up having international status if
not broadly used, but consider that it may be used in all personal health
records used by elite cyclists.

The ideal is to make archetypes as broadly used as possible to facilitate
interoperability, but if they need to be develope for a specific or one-off
purpose, or for use within just one healthcare facility, then that is still
a valid use of archetypes albeit at the cost of some 'shareability'.

Heather

12

Andrew,

I will just point out a very basic fact in openEHR: the idea is not that
the current RM will do everything for the future; indeed, the RM will
slowly be added to with new data types and other useful primitive
structures etc to allow proper representation of 'recorded information
about a subject' (which is the scope of openEHR). I can well imagine
adding a 3 or 4 coordinate data type to represent GPS position.

If what you want to record is no longer even 'recorded information about
a subject' - e.g. aircraft body parts, then you need a different base
reference model, and you can archetype that....

- thomas

Andrew Patterson wrote:

13

Think about it as an opportunity. Sports Medicine is a very data rich environment.

This provides a very good testing ground for archetypes.

Gordon Tomes
Acute Care Division
Department of Health and Ageing (MDP 63)
PO Box 9848, Canberra ACT 2601
Ph 02 6289 5081 | Mobile 0423 024 922 | Fax 02 6289 7630

**"Heather Leslie" <heather.leslie@oceaninformatics.biz>** Sent by: openehr-technical-bounces@openehr.org

02/07/2007 08:04 PM

Please respond to
For openEHR technical discussions <openehr-technical@openehr.org>

To

“‘For openEHR technical discussions’” <openehr-technical@openehr.org>

cc

Subject

RE: modeling non-medical data [No Protective Marking]



|

  • | - |

I think it is wrong to be too 'pure' about what constitutes health
information. In this situation a heart rate archetype can be specialized
for this very specific purpose, and the cadence etc is important part of the
context of the observation, being part of a very detailed record relating to
exertion, and should be recorded in their health record as state data. I
don't see any particular problem with GPS being recorded here as well if it
fulfils a specific need.

This particular specialization may not end up having international status if
not broadly used, but consider that it may be used in all personal health
records used by elite cyclists.

The ideal is to make archetypes as broadly used as possible to facilitate
interoperability, but if they need to be develope for a specific or one-off
purpose, or for use within just one healthcare facility, then that is still
a valid use of archetypes albeit at the cost of some 'shareability'.

Heather

_______________________________________________
Dr Heather Leslie
Director, Senior Clinical Consultant
Ocean Informatics Pty Ltd
M +61 418 966 670 (in Australia)
M +44 7722 064 546 (in UK)
Skype - heather.leslie.oi

>-----Original Message-----
>From: openehr-technical-bounces@openehr.org [mailto:openehr-technical-
>bounces@openehr.org] On Behalf Of Andrew Patterson
>Sent: Monday, 2 July 2007 7:31 PM
>To: For openEHR technical discussions
>Subject: Re: modeling non-medical data
>
>> But the purpose then would not be to collect data for healthcare
>> purposes - i.e. this post is about the use of open EHR standards and
>> technologies for to answer a scientific question?
>
>Well, this is dilemna I guess - it's data about a person, created
>from a physical activity - but it's not a clinical observation. It
>is being used for healthcare purposes, albeit with very
>specialised goals - the data is used to set safe training exertion
>ranges etc for the atheletes.
>
>I may end up doing it with an ad-hoc system - I just thought it
>would be interesting to consider how it might be done using
>openehr.
>
>Andrew
>_______________________________________________
>openEHR-technical mailing list
>openEHR-technical@openehr.org
>http://lists.chime.ucl.ac.uk/mailman/listinfo/openehr-technical
>
>__________ NOD32 2368 (20070701) Information __________
>
>This message was checked by NOD32 antivirus system.
>http://www.eset.com

_______________________________________________
openEHR-technical mailing list
openEHR-technical@openehr.org
http://lists.chime.ucl.ac.uk/mailman/listinfo/openehr-technical