10 Dutch hospitals choose for EPD from Google or Microsoft

This announced yesterday, together they form 20% of the market. There is
a lot of press attention, even from non-traditional ICT/healthcare press.
Seems a important move, after seven years designing, hospiutals choose
something completely different.

Some links for the Dutch readers. (others can translate them with Google
translate)
http://www.automatiseringgids.nl/markt-monitor/nederland/2008/37/ziekenhuizen%20pleiten%20voor%20google%20in%20plaats%20van%20epd.aspx
http://webwereld.nl/articles/52689/ziekenhuizen-willen-google-en-microsoft-als-epd.html
http://www.qure.nl/modules.php?op=modload&name=News&file=article&sid=3178
http://www.computable.nl/artikel/ict_topics/overheid/2700050/1277202/ziekenhuizen-zien-epd-niet-zitten.html

Bert

Hi Bert,

Thanks for sharing the news. Google Health and Microsoft Health Vault are certainly catching momentum (similar posts on other threads) and we, openEHR as a community should react to it.

For example, it will be very interesting to explore the possibilities of import/export data in/out Google Health and Microsoft Health Vault using openEHR/13606 archetypes. Google Health seems to use a subset of the CCR (Continuity of Care Record) to transport data, a quick way to achieve archetype-based data transfer would be a CCR template with archetypes from the openEHR knowledge repository. Later if possible, we should try to convince Google and Microsoft to add support for openEHR/13606 the APIs.

Cheers,
Rong

Rong Chen schreef:

Hi Bert,

Thanks for sharing the news. Google Health and Microsoft Health Vault are certainly catching momentum (similar posts on other threads) and we, openEHR as a community should react to it.

For example, it will be very interesting to explore the possibilities of import/export data in/out Google Health and Microsoft Health Vault using openEHR/13606 archetypes. Google Health seems to use a subset of the CCR (Continuity of Care Record) to transport data, a quick way to achieve archetype-based data transfer would be a CCR template with archetypes from the openEHR knowledge repository. Later if possible, we should try to convince Google and Microsoft to add support for openEHR/13606 the APIs.

One problem is, I am not aware of a (published) API-set (or did I miss something?), I am writing my own, and I guess others will do also. An important reason is the absense of a standarized API-set. One problem is that OpenEhr is not ready yet.

Bert

Rong Chen schreef:

Hi Bert,

Thanks for sharing the news. Google Health and Microsoft Health Vault are certainly catching momentum (similar posts on other threads) and we, openEHR as a community should react to it.

For example, it will be very interesting to explore the possibilities of import/export data in/out Google Health and Microsoft Health Vault using openEHR/13606 archetypes. Google Health seems to use a subset of the CCR (Continuity of Care Record) to transport data, a quick way to achieve archetype-based data transfer would be a CCR template with archetypes from the openEHR knowledge repository. Later if possible, we should try to convince Google and Microsoft to add support for openEHR/13606 the APIs.

One problem is, I am not aware of a (published) API-set (or did I miss something?), I am writing my own, and I guess others will do also. An important reason is the absense of a standarized API-set. One problem is that OpenEhr is not ready yet.

In this case, we use the API of Google Health Data to send or retrieve data so the openEHR EHR Service API isn’t needed. Well, this really isn’t the best mailing list for this type of discussion, let’s continue this on technical or implementer list.

Cheers,
Rong

In this case, we use the API of Google Health Data to send or retrieve data

so the openEHR EHR Service API isn’t needed. Well, this really isn’t the

best mailing list for this type of discussion, let’s continue this on

technical or implementer list.

I think you are right, this must be done, because Google and Microsoft will be very dominant.

But i wouldn’t surprise me that this i going to be a hell of a job.

Maybe it is a good idea to do this on open source base, and organize things, so we are not all doing the same thing ten times, and in the end being incompatible to each other..

The problem is, I don’t have much time, but I feel the urge to look for opportunities. I can, however, spent some (f.e. four) hours every week on this subject. I hope others can do that also.

My suggestion is that we all, who wants to be involved, start reading, and next Monday, in the evening, or Tuesday, have some idea what a good approach to this matter can be.

Bert

In this case, we use the API of Google Health Data to send or retrieve data

so the openEHR EHR Service API isn’t needed. Well, this really isn’t the

best mailing list for this type of discussion, let’s continue this on

technical or implementer list.

I think you are right, this must be done, because Google and Microsoft will be very dominant.

But i wouldn’t surprise me that this i going to be a hell of a job.

Maybe it is a good idea to do this on open source base, and organize things, so we are not all doing the same thing ten times, and in the end being incompatible to each other..

The problem is, I don’t have much time, but I feel the urge to look for opportunities. I can, however, spent some (f.e. four) hours every week on this subject. I hope others can do that also.

My suggestion is that we all, who wants to be involved, start reading, and next Monday, in the evening, or Tuesday, have some idea what a good approach to this matter can be.

Bert

Exactly! I think an open source implementation makes lots of sense here. So if we do it once, all openEHR compliant system will be able to transfer data with Google Health. =)

I just created a wiki page on this topic. It’s just a start. Feel free to contribute or comment. Let’s talk about this on the Java list from next week.

Cheers,
Rong

Ahem!

Something that existed before either Google or MS offerings as PING. Now called IndivoHealth.
http://www.indivohealth.org/

software LGPL license Indivo CHIP are trademarks of Children’s Hospital Boston.

Hi Bert,

Thanks for sharing the news. Google Health and Microsoft Health Vault
are certainly catching momentum (similar posts on other threads) and
we, openEHR as a community should react to it.

...

if possible, we should try to convince Google and Microsoft to add
support for openEHR/13606 the APIs.

Yes: if you can't beat them, join them! And infect them from the inside
:slight_smile: An openEHR member with stature could contact one (or more) of the
members of Googles Health Advisory Council:
http://www.google.com/intl/en-US/health/about/ghac.html
and arrange for becoming a member him/herself (Really! I mean, even
Wal-Mart is represented)

Roger

Ahem!

Something that existed before either Google or MS offerings as PING.
Now called IndivoHealth.
http://www.indivohealth.org/
software LGPL license Indivo CHIP are trademarks of Children's
Hospital Boston.

Probably there are hundreds of such offerings around the world, but none
of them had such marketing power as MS and Google have. When they put
their weight behind something, it's got momentum!

Roger

I just created a wiki page on this topic. It's just a start.

Good start!

Feel free to contribute or comment. Let's talk about this on the Java
list from next week.

Cheers,
Rong

If it is to be platform independent, I'd prefer to keep the discussion
on this list.

Roger

For what it's worth, we will publish our (Ocean) virtual EHR service API
as a proposal for openEHR very soon. I encourage others with service
interfaces to do the same in the interests of getting a standardised
virtual EHR API agreed in the next 6-12 months.

- thomas beale

All information below is published in the Dutch press, specially, the always good informed website on health ICT: Qure (http://www.qure.nl).
Normally, I don’t repeat press-messages on this or any other mailinglist, but this looks very important to many of us, who are working on a OpenEhr-implementation. I hope OpenEhr related press will occur so frequently that we don’t need to know. Anyhow, a business-succes for Openehr related software is in my opinion a success for all of us. It means that OpenEhr will be something management have to think/talk about

The latest update on this thread is, that the MCA (medical centre alkmaar) will run, as first hospital, software based on the European EPD standard (CEN 13606). This is also said by Hans Kedzierski (the person who announced that 10 hospitals are going to use the software from Google or Microsoft), member of the board of the MCA, and is published last friday
This latest announcement will be explained on a public meeting in october. Interesting are the companies who are working together on this: HP, de 13-Groep, ERC, Unusual Visions, Carelliance/Medical Insight, Technicolor, ASP4All, Eurofiber en Priority Telecom. Special interesting for us is ERC (http://www.e-recordcompany.eu/), reseller of Ocean Software. ERC seems a brand new company, its website is not (yet) fully functional (need some polishing).

Does this mean that the MCA will be using the two systems simultaneous?
Does this mean that there will be data-migration software between OpenEhr and “Google or Microsoft” (always mentioned together by Kedzierski)?
And if, how does this reflect on us? Will Google create an OpenEHR-data-exchange API?

More on this you can read in on the news-site Qure:
http://www.qure.nl/modules.php?op=modload&name=News&file=article&sid=index.php?name=News&file=article&sid=3179&mode=thread&order=0&thold=0
(it is not a free news-service)

Hi Berth,
Please let me add my set of questions to yours, for I have not been able to figure out the overall scheme of the discussed setup. Maybe its my fault, and I apologize in advance if I’ve missed the answers in the discussion or on the web.
What is the extend of Google’s or Microsoft’s offer to these hospitals? Both MS and Google give me the impression that their offer is about personal healthcare records. The extend of functionality they will offer is important, because it can either be a complementary functionality that will allow the patient to use his/her initiative to transfer data to allowed parties, like the next GP or hospital, or it can be a complete alternative to whole idea of national healthcare information systems, which is unlikely in my opinion.
What about the legal consequences? For example: I can remember being told my multiple sources that it is not legal to store healthcare data of a UK citizen outside of UK. Will the mentioned solution be located in USA? Are there similar legal constraints for discussed setup?

Cheers
Seref

Hi Seref,

You said:
"What about the legal consequences? For example: I can remember being
told my multiple sources that it is not legal to store healthcare data
of a UK citizen outside of UK."

I am not aware of any law restricting the storage of healthcare data
outwith the UK. I suspect there may be some confusion around the UK
Data Protection legislation which prohibits 'personal' data being held
about a subject, without the organisation holding the data being
registered. I doubt if the same could be held to apply to a patient
wishing to store their own data in an overseas hosted environment.

I can understand that many healthcare providers and national
organisations are attracted by the Google/HealthVault model as by
putting the patient in charge of the visibility of a potentially
widely accessible health record, it relieves some of the difficult
privacy and confidentiality issues. It is also fairly easy for the
various system vendors to populate the very limited set of CCR
clinical data classes via well-documented APIs.

The only problem I see is that this is being touted as a panacea for
all the well-known difficulties of implementing the EHR and ignores
the the necessity for a useful EHR to be 'maintained' so that it
reflects current knowledge, rather than just being a historical list
of observations and evaluations which may conflict and will certainly
not allow accurate workflow and scheduling support.

Ian

Dr Ian McNicoll
office / fax +44(0)141 560 4657
mobile +44 (0)775 209 7859
skype ianmcnicoll
ian@mcmi.co.uk

Clinical Analyst - Ocean Informatics ian.mcnicoll@oceaninformatics.com

Consultant - IRIS GP Accounts ian@gpacc.co.uk

Member of BCS Primary Health Care Specialist Group – www.phcsg.org

Hello Ian,
I’ll recheck to make sure about the status of the constraint I mentioned, and this is an argument I’d love to be wrong about.
IMHO Google and MS may be on the edge of falling into the well known trap for storing healthcare data: “lets build our own system”.
Especially Google’s other work like Google apps for cloud computing gives the impression that they are focused on scaling in the first place: no relational storage etc..
I’d love to see them using a richer layer for data input/output, and the following would be my first argument if I was given a chance to talk to them:
In case MS and Google omit the programmatic query aspects of the data they’ll be holding, they’ll simply push the cost of developing useful queries to users of that data. The users here must be perceived as (largely) GP software vendors, and secondary care solution providers, a.k.a hospital information system vendors. I do not consider the patient as the end user in terms of an information system, and this may sound a little bit strange. However, if you think about it, healthcare is an agent based service, where the resources are consumed by an agent (the doctor) on behalf of the patient. So, even if the patient can lift the burden of privacy management in the legal context from the shoulders of healthcare institutions, there is not much for him/her to do other than allowing access to his/her information for a particular user (another gp, HIS etc.)
So without a rich query layer in their systems, Google and MS will realize a system where a patient’s use of the system results in print outs, or a doctor using a web browser to access the data. Both scenarios will lead to doctor’s complaints: they’ll want things in a particular order, past medication in a particular format, etc… They’ll want this data in their usual system.For UK, they should be able to see this data based on CUI components for example. (my opinion again)
Again, without a rich query layer, the task of giving the doctors what they want is a cost item for the GP or HIS solution vendor. They’ll have to get a chunk of data with a blunt method, and perform operations on it since they do not have a nice query mechanism in the first place. This is programmer salary!
To have a larger market adoption and better integration with other vendors, these solutions by Google and MS will have to focus on strong interfacing features, and even though they do not use a standard backend, they’d better think about mechanisms better than CCR. Of course these are all my ideas, and all feedback/discussion is more than welcommed.

Cheers
Seref

Ian McNicoll schreef:

Hi Seref,

You said:
"What about the legal consequences? For example: I can remember being
told my multiple sources that it is not legal to store healthcare data
of a UK citizen outside of UK."

I am not aware of any law restricting the storage of healthcare data
outwith the UK. I suspect there may be some confusion around the UK
  

As I read somewhere, the EU had made a aggreement with Google to store
medical data for Europeans only in Europe.
Maybe I can find the link again.

Data Protection legislation which prohibits 'personal' data being held
about a subject, without the organisation holding the data being
registered. I doubt if the same could be held to apply to a patient
wishing to store their own data in an overseas hosted environment.

I can understand that many healthcare providers and national
organisations are attracted by the Google/HealthVault model as by
putting the patient in charge of the visibility of a potentially
widely accessible health record, it relieves some of the difficult
privacy and confidentiality issues. It is also fairly easy for the
various system vendors to populate the very limited set of CCR
clinical data classes via well-documented APIs.
  

Does someone have an opinion about the classes used by Microsoft?
You can find a link to the SDK here:
http://www.healthvault.com/Industry/program-overview.html

In the SDK is a Windows helpfile containing information about all classes.

The only problem I see is that this is being touted as a panacea for
all the well-known difficulties of implementing the EHR and ignores
the the necessity for a useful EHR to be 'maintained' so that it
reflects current knowledge, rather than just being a historical list
of observations and evaluations which may conflict and will certainly
not allow accurate workflow and scheduling support.
  

This seems true, but isn'yt it possible through the API's to add more
functionality?

Bert

Seref Arikan schreef:

Hi Berth,
Please let me add my set of questions to yours, for I have not been able to figure out the overall scheme of the discussed setup. Maybe its my fault, and I apologize in advance if I’ve missed the answers in the discussion or on the web.
What is the extend of Google’s or Microsoft’s offer to these hospitals? Both MS and Google give me the impression that their offer is about personal healthcare records. The extend of functionality they will offer is important, because it can either be a complementary functionality that will allow the patient to use his/her initiative to transfer data to allowed parties, like the next GP or hospital, or it can be a complete alternative to whole idea of national healthcare information systems, which is unlikely in my opinion.

Hi Seref, I am not sure what the extend is. I don’t know many details. I believe that there will be in Alkmaar a sort of cooperation between OpenEhr and Google or Microsoft. This, I conclude from combining information that comes to me.
The situation at this moment is: In the Netherlands there are (according my information) hardly hospital wide medical information-systems. Most hospital-systems are administrative.
This is because in a hospital clinicians often work in small business groups with all their own data-store. They pay hospitals for having a place to work, and other services.
Some services in hospital, like, f.e. laboratorium, are used by all clinicians, and have their data hospital-wide on a hospital-system.
(However, in academical (university) hospitals, this is completely different. Most hospitals in the Netherlands are not connected to universities.)

Now, the government does no longer accept this, and is in the opinion that there must be medical information systems, not only hospital-wide, but country-wide. They even set up a very expensive study- and design group NICTIZ, which in 7 years is to come with a product that maybe could be good 7 years ago, but it is not anymore today.
Like banks, insurances, shops, supermarket, even police and government, they offer Internet services. If there is a country-wide medical information system, the patient needs too his connectivities and responsibilities in this system. The NICTIZ system as far as is designed does not offer this. In fact it is a very old fashioned system, typical for the last century.

This is where Google and Microsoft come in. They fill in the information gap in many hospitals, they give responsibilities to the patient, the have Internet connectivity, they use most modern technical architecture (cloud), they offer an API for third parties to connect to, and large companies like these are well equiped for 24/7 99,999% uptime and save.

The only thing the Dutch government has to do is make service level-aggreements on where to store data, how to backup, which security-levels, etc.

The govenrment is not yet that far, they see it as a political blamage to give up NICTIZ, it costed hunreds of millions, in 7 years, and produced nothing that is yet working. The hospitals however, they don’t have the risc of blamage, but they have an information gap.
You know what they say in the Netherlands about sheep: if one is over the bridge, others will follow, something like that.

It could go fast.

What about the legal consequences? For example: I can remember being told my multiple sources that it is not legal to store healthcare data of a UK citizen outside of UK. Will the mentioned solution be located in USA? Are there similar legal constraints for discussed setup?

There is a lot to say about legal consequences, for example, copyright, who is the owner of the data but on the other hand, the patient will be responsible for what to store, and by whome visible, that must be a big relief. It maybe makes things in legal perspective simpel.
However, this is only what I think and read, and I am not an expert in this.

Bert