Principles of a Healthcare open platform


(gina.naughton) #21

That is exactly why we need Clinicians (and multiple other system users who depend on accurate data being available at the point of care and at every other step of the workflow) to become involved in the creation of the standards on which we will build EHR. I’m not a clinician but a systems manager/administrator and deal regularly with the fallout of lack of data integrity on systems. We need to fool proof the systems in so much as we can at data input level and make it easy and safe to access, verify and extract that same data when required. Humans will make errors and we need to acknowledge that and try to prevent recurrence when it does happen


(navin1976) #22

@beckywassall
When I started planning the modelling phase I realised that there was no standardised way of approaching this. Everyone just did it their own way. This would be very problematic for us, as the cancer model would be huge and would require multiple people working on it, all with a common way of working.

That is why I decided to document how we were going to do it, so that there would be a basic guide, and so that people could learn from our journey (ie mistakes!)
There will be bits which may not make sense to people outside our group, so please let me know, and I can explain / update the site with something more useful.

Thanks again!
Navin


(pablo.martin) #23

Indeed, without the voice of a mix of clinicians, each one with different needs you can not assess the complexity of the problem at hand, even less the paramount requirement of interoperability


(clive.spindley) #24

the stuff that is overloading the nhs is the admin data, that is where the focus should be, boring as that may

sound to the clinicians amongst us

... just need to nip down to pc world to sort something out (no big deal)

On 13 March 2016 at 21:47 navin1976 <discourse-system@openhealthhub.org> wrote:

navin1976

March 13

From what I understand, openEHR's forte is dealing with patient encounter and phenotypic data - not demographics or scheduling. I believe there are some workarounds for storing these other bits of data, but this is of uncertain benefit, and pragmatically these would likely be stored in other formats.

With the momentum that FHIR is gaining it would be a natural best-fit for openEHR to use FHIR to communicate with these other systems.

Anyway, that is my current understanding of the situation!

Navin


Visit Topic or reply to this email to respond

To stop receiving notifications for this particular topic, click here. To unsubscribe from these emails, change your user preferences


(navin1976) #25

It would be interesting to know what you define as the “admin data” Clive.
From my experience, specific to the cancer pathways, the “admin data” is often handled within the reference model data of openEHR. So by getting the clinical data correct, and having a good underlying reference model, we can tackle both together.

Again, this is purely my take on the situation, with admittedly limited technical knowhow.

Navin


(ian) #26

Thanks Navin,

I was going to ask exactly the same question.

@clive.spindley - I suspect your definition of ‘admin’ data might well be something that I would expect to appear, at least partially, in a patient record. As an example openEHR would expect to track the stages of a referral
Have a look at http://openehr.org/ckm/#showArchetype_1013.1.2374_MINDMAP , in particular the Pathway branch of the mindmap.

Formal scheduling is out-of-scope but would be tied to the patient record via a workflow_id reference.

I wholly agree that we do not focus on the ability of IT to improve the patient journey. If nothing else, knowing what the heck is going on and any point in that journey, would make a huge difference to both the patient and their clinicians.

Ian


(ian) #27

Hi Kevin.

As ever … it depends … :slight_smile:

There are a few options.

  1. AQL is sufficiently rich that you could carry out reporting style queries directly on the data, and it would not take long for people familiar with SQL to get up to speed. What they would find confusing is the complexity of the underlying operational data (true of any front-line implementation), although of course corresponding benefits from allowing much more nuanced reporting. There would also be inevitable performance hits from running reporting on an operational system.

  2. You could mirror the openEHR data to a second service which gets around the performance issue, and allows the use of an openEHR CDR optimised for analytics-type queries.

  3. Good old ETL out to a traditional RDBMS. Everyone is in their comfort zone but you are still at the mercy of someone understanding the optimal simplifications as the data is simplified on extract/transform. I would suggest that it is this simplification that is significant, rather than familiarity with SQL vs. AQL but with simplification comes the downside that the reporting dataset is limited.

Ian


(ian) #28

and just as importantly we need experienced clinical informaticians (some of whom might have a technical background) who are able to mediate these conversations within the bigger picture of the complexity of health information. We really have a dearth of practical informatics expertise and almost no support for this as a career.

Many Informatics courses have a somewhat theoretical approach and fail to get inside the practicalities of implementation. I would love to see clinical informatics trainees embedded inside some real-world, commercial projects, learning their trade.

Whilst ‘ordinary’ clinicians need to have their voices clearly heard, this is is not an area where it is easy for someone like @Becky to easily get up to speed and still do the day job.


(Kev Mayfield) #29

@Ian,

I think I’ve gone off on a tangent at a much earlier stage than that. Misreading terms which meant regional interop infrastructure to me (MPI and XDS), I’m assuming Patient Administration System (PAS) and Document Management systems is an alternate description. OpenEHR was not going to be the all consuming single regional database.

Something like most GP’s have (TPP/EMIS/INPS for openEHR and Docman for Doc Management)?


(Kev Mayfield) #30

p.s. the data warehouse is for national reporting (is this ‘admin data’ - I would call it that). So doesn’t need to hold clinical data :slight_smile:


(navin1976) #31

This is still clinical data. We prefer the term “operational” for the data that is used for clinical care.
This operational data feeds analytic systems, be they registries, business intelligence or research platforms (which in turn can feed back into the operational systems):

That is the whole point of the opencancer community - see:
http://opencancer.net/index.php/2015/12/18/opencancer-is-born/

Navin


(clive.spindley) #32

i might be naive but i think things aren't going to bad, re:clinicial data, in fact they are bloody good!

i think that is 'cause historically clinicians have ruled the roost, nothing wrong with that, but the admin data has

fallen behind

you are right, what is required is a model that can handle both, model that is efficient and all encompassing

so clinicians do have to do their own thing

wot do i mean bi admin data?

dates and times, so that durations can be derived, i.e. the when

(length of hospital stay is discharge d/t - admission d/t, that, to me, is admin stuff)

costs at a patient level, so that accounting can be accurate and not, finger in the air job (be honest !!!) i.e. how much

(efficient use of the money is becoming much more critical)

locations, organisations, wards i.e. physical stuff, the where (integartion)

lots more, i will be providing a link to my data model shortly, but yes, it can handle clinical indicators

(albeit, fairly basic, duration of chemo = end of chemo d/t - start of chemo d/t)

On 14 March 2016 at 11:42 navin1976 <discourse-system@openhealthhub.org> wrote:

navin1976

March 14

It would be interesting to know what you define as the "admin data" Clive.

Again, this is purely my take on the situation, with admittedly limited technical knowhow.

Navin

Visit Topic or reply to this email to respond


In Reply To

clive.spindley

March 14the stuff that is overloading the nhs is the admin data, that is where the focus should be, boring as that may sound to the clinicians amongst us … just need to nip down to pc world to sort something out (no big deal) On 13 March 2016 at 21:47 navin1976 <discourse-system@openhealthhub.org> wrote: …


Visit Topic or reply to this email to respond

To stop receiving notifications for this particular topic, click here. To unsubscribe from these emails, change your user preferences


(clive.spindley) #33

'tell you what, let's break down the walls, clinical, operational, financial, cancer, national reporting (aggregated), admin ... blah blah

let's just call it PATIENT health data (or PHD if you are in to that sort of thing), after all, it is the patients data and the NHS are the guardians of that data, and the clinicians need the data to help the patient, but it is the PATIENT's data on the PATIENT's pathways

it's the PATIENTs choice who, if anyone, they want to share the data with (i am sure they will want to share it with the clinicians as they TRUST the clinicians and know the clinicians will not pass the data on and are trying to help them not harm them, they have sworn an oath) - they may choose to share the anonymized or pseudonamized data outside the NHS in order to help others, but that is their (the PATIENTs) choice, no one else's, as it is their (the PATIENTs) data on the PATIENT's pathways ...

you may have gathered now, i am non clinical, my specialty is PATIENT health data (and how to turn it into helpful information)

On 14 March 2016 at 19:23 "clive.spindley" <discourse-system@openhealthhub.org> wrote:

clive.spindley

March 14

i might be naive but i think things aren't going to bad, re:clinicial data, in fact they are bloody good!

i think that is 'cause historically clinicians have ruled the roost, nothing wrong with that, but the admin data has

fallen behind

you are right, what is required is a model that can handle both, model that is efficient and all encompassing

so clinicians do have to do their own thing

wot do i mean bi admin data?

dates and times, so that durations can be derived, i.e. the when

(length of hospital stay is discharge d/t - admission d/t, that, to me, is admin stuff)

costs at a patient level, so that accounting can be accurate and not, finger in the air job (be honest !!!) i.e. how much

(efficient use of the money is becoming much more critical)

locations, organisations, wards i.e. physical stuff, the where (integartion)

lots more, i will be providing a link to my data model shortly, but yes, it can handle clinical indicators

(albeit, fairly basic, duration of chemo = end of chemo d/t - start of chemo d/t)

On 14 March 2016 at 11:42 navin1976 <discourse-system@openhealthhub.org> wrote:

#3b5998; ; font-size: 13px; font-family: 'lucida grande',tahoma,verdana,arial,sans-serif;" href="http://mandrillapp.com/track/click/30623946/www.openhealthhub.org?p=eyJzIjoicTAzMzNiU0ZZOVd2dS1tMUxNd3A5UkUydjE0IiwidiI6MSwicCI6IntcInVcIjozMDYyMzk0NixcInZcIjoxLFwidXJsXCI6XCJodHRwOlxcXC9cXFwvd3d3Lm9wZW5oZWFsdGhodWIub3JnXFxcL3VzZXJzXFxcL25hdmluMTk3NlwiLFwiaWRcIjpcIjVmMDM0NDQ1ZDEyMzQzZjZiY2E4ZmYyZGRmNWQ4NGNiXCIsXCJ1cmxfaWRzXCI6W1wiZTcxMGY4Y2U3ZGEzNTc3OGI4MmEzMGIzMTcxNDljNzQzMTk4NjYzMVwiXX0ifQ" target="_blank">navin1976

March 14

It would be interesting to know what you define as the "admin data" Clive.

Again, this is purely my take on the situation, with admittedly limited technical knowhow.

Navin

Visit Topic or reply to this email to respond


In Reply To

clive.spindley

March 14the stuff that is overloading the nhs is the admin data, that is where the focus should be, boring as that may sound to the clinicians amongst us ... just need to nip down to pc world to sort something out (no big deal) On 13 March 2016 at 21:47 navin1976 <discourse-system@openhealthhub.org> wrote: …


#ddd; height: 1px; border: 1px; ; background-color: #ddd;">

#006699; ; color: #666;" href="http://mandrillapp.com/track/click/30623946/www.openhealthhub.org?p=eyJzIjoiSHNDRlJFb2NvREdBbVFqLUtXN3owMEg4QTFVIiwidiI6MSwicCI6IntcInVcIjozMDYyMzk0NixcInZcIjoxLFwidXJsXCI6XCJodHRwOlxcXC9cXFwvd3d3Lm9wZW5oZWFsdGhodWIub3JnXFxcL3RcXFwvcHJpbmNpcGxlcy1vZi1hLWhlYWx0aGNhcmUtb3Blbi1wbGF0Zm9ybVxcXC8xNzRcXFwvMjVcIixcImlkXCI6XCI1ZjAzNDQ0NWQxMjM0M2Y2YmNhOGZmMmRkZjVkODRjYlwiLFwidXJsX2lkc1wiOltcImY1MDMyMTk0NjNmMTlkODQyMzU5YzlmOTM3YzliY2I1MWM4NDhjY2NcIl19In0">Visit Topic or reply to this email to respond

To stop receiving notifications for this particular topic, click here. To unsubscribe from these emails, change your user preferences

Visit Topic or reply to this email to respond


In Reply To

navin1976

March 14It would be interesting to know what you define as the “admin data” Clive. From my experience, specific to the cancer pathways, the “admin data” is often handled within the reference model data of openEHR. So by getting the clinical data correct, and having a good underlying reference model, we can …


Visit Topic or reply to this email to respond

To stop receiving notifications for this particular topic, click here. To unsubscribe from these emails, change your user preferences


(navin1976) #34

I might be missing something here Clive. I tried the Google drive link you posted in the pathways thread, but the folder appeared to be empty (may be my error).

But are you saying you have built your own data models? Are they aligned to any particular ontoligies or terminologies? Have they been sense-checked by any users of the systems (eg admin staff, patients, nurses, doctors etc), and do they tally with the standards produced by professional bodies such as the Association of Royal Medical Colleges.

I agree that sometimes clinicians can ask for some odd things, but they still need to be engaged. After all, this stuff gets quite specific. Ask me about neurosurgery data needs and I can have a guess, but after the basic few components, I will be lost.

That is the beauty of the openehr clinical knowledge manager, allowing multiple people of different backgrounds to contribute to the process, rather than any single person alone.

Any open model has to have that widespread input to succeed.

Anyway, I may have got the wrong end of the stick.

Navin


(Pablo Pazos Gutiérrez) #35

Coming late, but did you check GitHub - ppazos/cabolabs-ehrserver: Service-oriented openEHR repository for clinical data with composition commit, query and retrieve capabilities.?