Monday, June 30, 2014

Weekly Australian Health IT Links – 30th June, 2014.

Here are a few I have come across the last week or so.
Note: Each link is followed by a title and a few paragraphs. For the full article click on the link above title of the article. Note also that full access to some links may require site registration or subscription payment.

General Comment

Clearly the big news for the week was the Departure of David Gonski as the Chair of NEHTA and the ascension of Dr Steve Hambleton to the role. I have provided some thoughts on this earlier.
Also interesting is the departure of Jane Halton from DoH to the inner circle of Government, but not, however, as close to the centre as she may have desired.
It is also interesting that it seems the GP Co-Payments seem to have a technical issue that may be pretty expensive to fix!
All in all quite a lively week.
-----

Gonski steps down as NeHTA chair, replaced by Hambleton

High profile corporate and education leader David Gonski has stepped down from his role as the chairman of the National eHealth Transition Authority and will be replaced by the immediate past president of the Australian Medical Association (AMA), Dr Steve Hambleton.
The changing of the guard comes as the cross-government body tasked with making a national electronic health and medical records scheme a functional reality faces a pivotal year after its funding was extended for just one year in the federal Budget.
The departure of Mr Gonski was widely anticipated as he is ineligible to serve a third term as chairman under NeHTA’s constitution.
Mr Gonski, whose name has since become synonymous with the ambitious education reforms of the former Labor Federal government, became the NeHTA’s chairman in 2008 and is widely credited with keeping the massive and frequently challenged project alive through his quiet but formidable style of diplomacy.
-----

Privacy fears curb e-health’s growth

June 25, 2014
Mark Eggleton
In a twist on the old highwayman demand of “Your money or your life” we finally have an answer when it comes to e-health. Our personal finances win while our health takes a back seat.
Right now most Australians when they want to get a snapshot of their financial situation can go online and find up-to-the-minute information on their bank balance and outstanding debts. We’re pretty comfortable with the level of security afforded our financial details and even happy to give out further details if we’re keen on purchasing goods or services. Unfortunately, we’re a little leery about having our health records available online beyond what’s stored in a computer on our GPs desk.
Security of data was one of the major focuses of the recent Big Data in Healthcare roundtable held by The Australian Financial Review in partnership with GE in Sydney with most participants agreeing it was an issue.
-----

Tech giants launch new e-health network

Christina Farr and Thomas Escritt
Dutch healthcare and lighting company Philips said it was teaming up with Salesforce, one of the biggest cloud-computing companies in the United States, to offer online management of chronic diseases.
The venture will involve a cloud-based software platform that will take data fed from networked medical devices in homes and hospitals to allow nurses and doctors to monitor the health of hundreds of patients simultaneously.
The companies said they would launch two new medical applications, Philips eCareCoordinator and Philips eCare Companion, later in the summer as part of the partnership.
Philips’ chief executive Frans van Houten told journalists in Amsterdam the services would make it easier and cheaper to monitor the health of patients.
-----

Verbal nursing handovers go digital

Dr Leif Hanlen is the Health Team Director at NICTA, Australia’s Information Communications Technology (ICT) Research Centre of Excellence in Canberra, where he is currently involved in several projects that focus on large scale ehealth data analysis – and have the potential to improve patient outcomes while also reducing workloads for health professionals.
A recently completed project used the text in hospital radiology reports to identify an infection, invasive aspergillosis, threatening transplant and cancer patients.
“We’re looking at text processing services that help the clinical work flow, rather than trying  to develop a brand new text or health analytics box that adds to the clinical burden,” he says.
One key project involves recording and classifying the clinical data transferred in a typical nursing hand-over to improve accuracy of information and reduce the administrative burden.
-----

Jane Halton leaves Health for Finance, Treasury still open

Prime Minster Tony Abbott has announced another round of musical chairs at for Canberra’s public service chiefs, but put off naming a replacement for outgoing Treasury head Martin Parkinson.
High profile Secretary of the Department of Health, Jane Halton will take on the role of nailing down government expenses as the head of the Department of Finance as a replacement for David Tune who has opted to retire from the Australian Public Service and will finish-up his term at the end of this week.
The move by Ms Halton to Finance, for a five year appointment, is likely to fuel continued speculation that she is a likely successor to incumbent APS chief and Secretary of the Department of Prime Minister and Cabinet, Dr Ian Watt.
-----

System to track GP visits ‘a big ask’

Fran Foo

Technology Reporter
Sydney
TAXPAYERS will have to pay “hundreds of millions of dollars” to build a centralised database for 25 million people that would update in real time to cater for the Abbott government’s proposed $7 medical co-payment scheme, IT experts say.
The government has proposed that people pay $7 each time they visit a GP, get an X-ray or a blood test from July next year. A patient who visits a doctor and needs a pathology test and an X-ray will be slugged with $21 in upfront fees.
The $7 fee is applicable to everyone except concession card holders and children under 16 who will pay for the first 10 services combined.
The controversial plan has drawn the ire of consumers and many in the healthcare fraternity who say it marks the demise of universal access to healthcare in Australia.
-----

New breast cancer tests finds more cancers than traditional mammograms

Date June 26, 2014

Amy Corderoy

Health Editor, Sydney Morning Herald

Adding three-dimensional mammograms to normal mammograms can catch breast cancer earlier, a new study has found.
But experts warn that the study, while promising, does not show 3D mammograms provide any long-term benefits over traditional screening, which is provided free to all women aged over 40 in Australia.
The study analysed data from more than 450,000 examinations, some of which were only done with mammograms and others that used the combination test, which captures multiple images of breast tissue that are built up to provide a 3D image.
The combined test caught 41 per cent more cancers than traditional mammography alone, according to the research, which was published in the Journal of the American Medical Association and funded by Hologic, which produces 3D mammography machines.
-----

Fitness bracelets ‘a dying business’, says Misfit CEO

  • Katherine Rosman
  • The Wall Street Journal
  • June 25, 2014 7:02AM
IN another sign that the shine is coming off the much-ballyhooed wearable hardware market, health-tracking device maker Misfit is announcing today that its software and its iOS app will be compatible with the smartwatch from competitor Pebble.
Pebble says the partnership is part of its longstanding strategy to make its watch a dynamic tool for those who like to monitor fitness and health metrics. But for Misfit, the move is partially influenced by the anticipation of Apple’s rumoured plan to release a smartwatch later this year.
The Shine is sold in Apple stores, which makes the device particularly vulnerable.
“If you buy one of these [Apple] devices you’re not going to buy an activity tracker,” says Sonny Vu, Misfit’s chief executive, adding that he predicts Android-powered watches will undo the fitness bracelet market. “Why should we cling on to a dying a business?”
-----

Data from wearable devices could help

  • Elizabeth Dwoskin and Joseph Walker
  • The Wall Street Journal
  • June 26, 2014 12:00AM
MANY runners and fitness fanatics have been quick to embrace wearable wireless tracking devices for measuring physical activity and calories burned. Now, a growing number of physicians are formally studying whether such “wearables” can improve patients’ health by spurring people to get moving.
Amy Wheeler, a doctor at Massachusetts General Hospital, hopes that wireless tracking devices can help motivate her obese patients to do what they haven’t been able to on their own: lose weight.
Last year, Dr Wheeler was one of a group of US doctors who gave FitLinxx pedometers to 126 patients with Type 2 diabetes, often related to poor diet and excess weight.
The pedometers tracked how many steps the patients took and linked to a software program that calculated whether they met their exercise goals. Based on their progress, data from their electronic medical records and whether it was sunny or rainy, patients would receive motivational tips via text message.
-----

Ambulance Victoria rushing data to hospital as well

Date June 24, 2014

Trevor Clarke

Ambulance Victoria has joined the big data movement by establishing a real time data exchange system to provide hospitals greater visibility of incoming patients' needs.
The organisation has been working with four hospital emergency departments in the state for the past two months to trial the new arrivals system.
“That is a real time reporting stack where we take a half million events, including GPS coordinates, and present that out to the hospitals on a large monitor in the [emergency department] so they know, for example, that an ambulance is arriving and it is a priority one and the estimated time of arrival is 10 minutes," Ambulance Victoria manager of enterprise architecture, John Dousset, told IT Pro.
The data comes from Emergency Services Telecommunications Authority (ESTA), which runs the 000 number. This feed provides 500,000 transactions or “events” every day and includes information such as when an ambulance is dispatched and arrives, along with GPS coordinates of ambulances in motion (every 300 metres) or stationary (every 15 minutes). The information is fed to emergency departments at Monash, Austin, the Northern and St Vincent hospitals. 
-----

Robot ‘legs’ help people walk

  • AP
  • June 28, 2014 12:00AM
US health regul­ators have approved a first-of-a-kind set of robotic leg braces that can help some disabled people walk again.
The ReWalk system functions like an exoskeleton for people paralysed from the waist down, allowing them to stand and walk with assistance from a caretaker.
It consists of leg braces with motion sensors and motorised joints that respond to subtle chan­ges in upper-body movement and shifts in balance.
-----

Apple, Google, Samsung vie to bring blood sugar apps to wearables

Date June 24, 2014 - 8:54AM

Christina Farr

For decades, medical technology firms have searched for ways to let diabetics check blood sugar easily, with scant success. Now, the world's largest mobile technology firms are getting in on the act.
Apple, Samsung and Google, searching for applications that could turn nascent wearable technology like smartwatches and bracelets from curiosities into must-have items, have all set their sights on monitoring blood sugar, several people familiar with the plans say.
These firms are variously hiring medical scientists and engineers, asking US regulators about oversight and developing glucose-measuring features in future wearable devices, the sources said.
The first round of technology may be limited, but eventually the companies could compete in a global blood-sugar tracking market worth over $12 billion by 2017, according to research firm GlobalData.
-----

Governments' IT projects needn't cost the world. Here's how we can do better

Australia has ambitious targets to improve the relationship between citizens and the government: all correspondence will be able to be conducted online by 2017
The man responsible for the UK government’s technology, Liam Maxwell, walks around with a very simple motto stickered onto his smartphone and Macbook: “What is the user need?
Maxwell is Her Majesty’s government’s chief technology officer but his outlook, attitude and clothes he wears are far closer to Steve Jobs than to Sir Humphrey.
His job as part of the efficiency reform group – created in 2010 when the UK government was facing its largest deficit since the second world war – is to equip government departments with the right technology to deliver great digital services and to cut IT spending. But it’s more accurate to say his real mission is closer to reimagining the role of government and its daily relationship with citizens.
-----

In your face: booming biometrics puts security ahead of privacy

Date June 25, 2014

Natasha Singer

The row over covert use of computer face recognition technology was kicked off in 2001 and there's no end in sight, writes Natasha Singer.
Who he? A New York supported at the infamous 2001 "snooper-bowl" final where police used embryonic face-recognition technology on the crowd - and launched a mighty row.
Joseph Atick watched the Ronald Reagan Building and International Trade Centre in Washington as if he owned the place. In a way, he did. He was an organiser of the event, a conference and trade show for the biometrics security industry. Perhaps more to the point, many of the wares on display, like an airport face-scanning checkpoint, could trace their lineage to his work.
A physicist, Atick is a pioneering entrepreneur of modern face recognition. Having helped advance fundamental face-matching technology in the 1990s, he went into business and promoted the systems to government agencies looking to identify criminals or prevent identity fraud.
‘‘We saved lives,’’ he says. ‘‘We have solved crimes.’’
-----

Astronauts to build giant telescope

  • The Times
  • June 23, 2014 12:00AM
SCIENTISTS are planning a space telescope so large it will have to be built by astronaut construction workers because no rocket is capable of lifting it from the Earth in one launch.
The Atlast, or advanced technology large-aperture space telescope, is being designed to take the first direct images of Earth-sized planets orbiting distant stars, allowing experts to analyse their atmospheres to see if alien life might have evolved on them.
The challenge of gathering light from such tiny and distant objects will, however, require a mirror measuring up to 16m in ­diameter — larger than anything of its kind yet built.
This means it would have to be assembled in space more than 1.5 million kilometres from Earth, four times further than the moon.
-----
Enjoy!
David.

17 comments:

Bernard Robertson-Dunn said...


Health has issued a couple of "consultation" papers and is looking for comment on the inclusion of Pathology and Diagnostic Imaging reports in the PCEHR

http://www.health.gov.au/internet/main/publishing.nsf/Content/eHealth

I've had a quick look at both papers and am stunned by what's not in the proposed models.

They have some very simplistic process flows but very little (that I can find) about the information/data that are involved in the process flows and in the reports themselves.

In Appendix C "Proposed metadata for PCEHR (pathology | diagnostic imaging) reports & ATP Messages" are some details about the proposed solutions but are vague and do not cover all the relevant information, neither do they cover the relationships between data.

There is also an amazing statement in both appendices: "The list below includes a new field ‘PCEHR Document ID‘- not previously discussed during consultation. It is proposed that the PCEHR Document ID is used as the unique identifier for a report across the integrated solution."

The PCEHR has been implemented and has been operational for two years. And it's only now that they have realised that there should be some way of uniquely identifying PCEHR documents.

So, they are now proposing to introduce a new data concept but have not provided any further information about its relationship with other data, its structure, how it is created, who creates it, etc etc...

Rule number 1 when building an Information System: Identify, define and understand the information - what it is, how it flows and all issues concerning its life cycle.

If these two documents are any indication of the thought that's gone into the PCEHR, then its even worse than I thought. And I thought it was pretty bad.

Someone please tell me that proper information/data descriptions actually exist and show me where I can find them....

Grahame Grieve said...

Bernard: "it's only now that they have realised that there should be some way of uniquely identifying PCEHR documents"

rofl. No, really? We got this system that doesn't know how to identify documents, but no one has noticed to now...

I don't think so. I don't have time to read the consultation papers, but they must be talking about introducing dependency on the identifier for some workflow as a change, because there (obviously) has been document identifiers from day 1.

You shouldn't read consultation papers as if they are technical specifications.. and I really think you could've worked this one out too

Bernard Robertson-Dunn said...

Grahame:

"I don't think so. I don't have time to read the consultation papers, but they must be talking about introducing dependency on the identifier for some workflow as a change, because there (obviously) has been document identifiers from day 1."

That's not the impression that the consultation paper gives and I'm not treating them as technical specifications, I'm assuming they are requirement definitions.

The details regarding "PCEHR Document ID" in the consultation papers (both of them) is:

Quote
PCEHR Document ID
In previous consultation it was proposed that the NATA ID, Local Report ID and Request ID could be combined to constitute a unique identifier for a report. However further advice received has indicated that this approach may not be feasible for the following reasons:

* Request ID will probably only be available for electronic requests;
* The Local Report ID may not always be globally unique; and
* Some pathology providers may use the same Local Report ID for the initial and subsequent versions of a report, i.e. Interim, Final, Corrected.

When a clinical document is made available in the PCEHR the author is responsible for assigning a globally unique document ID (PCEHR Document ID). It is proposed that the PCEHR Document ID is assigned at the same time that the report is generated so that it can be used as the key identifier for the document across the integrated solution.
End Quote.

Observations:
Whatever is in the existing system seems to be less than acceptable;

There may have been document identifiers from day one but they have not been unique;

There have been previous proposals to cobble together unique identifies but these are not workable;

The proposal is that "the PCEHR Document ID is assigned at the same time that the report is generated". How this is to happen is not made clear.

The statement "When a clinical document is made available in the PCEHR the author is responsible for assigning a globally unique document ID (PCEHR Document ID)." is ambiguous. Does this refer to the current situation or is it a proposal? The ATP proposal actually states that "When a clinical document is made available in the PCEHR the document author is responsible for assigning a globally unique document ID (PCEHR Document ID)." so it looks as though they are talking about a proposal.

I'd be interested in learning how a document author can generate a globally unique ID. Will the PCEHR system owners provide them with an algorithm that they can apply and generate a unique ID which is then attached to the report? Will the document author request a globally unique ID from the PCEHR system which is then attached to the report? Or what?

Grahame Grieve said...

ok, I can answer those questions:

"Whatever is in the existing system seems to be less than acceptable;"

No. This text you quoted was referring to a specific proposal for how actually globally unique identifiers might be derived from other existing identifiers for these specific documents (that do not yet exist in the PCEHR). The requirement for all documents to have a genuinely unique identifier is required by the international specifications.

"There may have been document identifiers from day one but they have not been unique;"

No, they always have been, and must be - the document identifier is the primary key in the PCEHR. ...are you familiar with primary keys? ;-)

"There have been previous proposals to cobble together unique identifies but these are not workable;"

yes, there has been a proposal for *these specific document types*, and I didn't think that the proposed scheme would work in all cases; so it has proved. The consequence is that the authoring systems must instead generate a unique identifier and store it (that derivation algorithm was an attempt to save them from storing a new identifier). (note: it's much easier if you can avoid this, because then a middleware engine can generate the documents without a change in the source system or a persistent store of it's own, which must be DR safely in sync with the source)

"The proposal is that "the PCEHR Document ID is assigned at the same time that the report is generated". How this is to happen is not made clear."

not sure what's not clear about this - the documents are assigned an id based on the PCEHR id that the document is generated (note my language changes a little, a matter of technical perspective - but it seems obvious to me)

"The statement "When a clinical document is made available in the PCEHR the author is responsible for assigning a globally unique document ID (PCEHR Document ID)." is ambiguous. Does this refer to the current situation or is it a proposal? "

It's the current situation for all existing documents/document types in the PCEHR (the author must assign a globally unique identifier), and a recognition that there is no simple workaround for the pathology case

"I'd be interested in learning how a document author can generate a globally unique ID."

UUID or OID. See my blog (http://www.healthintersections.com.au/?p=1002 - highly technical), or the HL7/ISO data type specifications, or read the NEHTA specifications which repeat all this.


"Will the PCEHR system owners provide them with an algorithm that they can apply and generate a unique ID which is then attached to the report?"

as above but this is not rocket science.

Bernard Robertson-Dunn said...

I think I'm getting more confused.

I get the impression that whatever is in the current system needs improving. The consultation papers and your comments on my observations seem to agree with this. If I've got this wrong then I'm completely lost.

So, there is a requirement to generate unique identifiers - something that currently doesn't happen.

I made the statement that "The proposal is that "the PCEHR Document ID is assigned at the same time that the report is generated". How this is to happen is not made clear."

Your response was "not sure what's not clear about this - the documents are assigned an id based on the PCEHR id that the document is generated (note my language changes a little, a matter of technical perspective - but it seems obvious to me)"

a) I made that statement because the consultation paper doesn't say how it's supposed to happen. You may have prior knowledge that is not available to at least some of us.

b) I can't make sense of your statement. What does "the PCEHR id that the document is generated" mean? For a start, what's a PCEHR id? The rest of the sentence doesn't appear to be normal English.

I asked the question "I'd be interested in learning how a document author can generate a globally unique ID." and your response was UUID or OID. I asked a process question and got references to things.

My concerns are not about IDs but about how they are generated. Care has to be taken when generating UUIDs in distributed systems. I would have thought that the consultation paper at least raise it as an issue.

Grahame Grieve said...

hmm. I'll try again

"there is a requirement to generate unique identifiers"

yes. it is required.

"something that currently doesn't happen."

It does happen. the unique identifier is the PCEHR primary key. So it *must* happen.

"You may have prior knowledge that is not available to at least some of us."

Perhaps, but it is available by reading the specs. I know that most people don't bother to read specifications, but if you want to comment in public, perhaps you should...

"b) I can't make sense of your statement. What does "the PCEHR id that the document is generated" mean? "

I missed a few words. Should have read "the PCEHR id that is generated when the report is created". Actually, the document has 2 unique identifiers in it - the instance identifier that is never used again, and the set identifier that is the same for later versions of the same document. The second must be generated with the report, the first must be generated with each individual document. I got my language scrambled while editing my sentence

"your response was UUID or OID. I asked a process question and got references to things."

Yeah, because you want me to rewrite entire stacks of specifications that are widely available on the net?

"Care has to be taken when generating UUIDs in distributed systems."

really? because that 1 in 10^50 chance isn't tolerable? If, by some distant chance, duplicate UUIDs were generated (see Wikipedia - http://en.wikipedia.org/wiki/Universally_unique_identifier#Random_UUID_probability_of_duplicates, or peruse https://www.google.com.au/webhp?sourceid=chrome-instant&ion=1&espv=2&ie=UTF-8#q=duplicate%20UUIDs%20generated&safe=off at your leisure), then the worst that will happen is that the system will reject a document when it is submitted as a duplicate.

And many systems use OIDs anyway, where there's no statistical chance involved at all.

"I would have thought that the consultation paper at least raise it as an issue."

I wouldn't. Even if the actual entropy is much lower than in theory - by many orders of magnitude - the risk is still minuscule, and the problems it can create are isolated to the author anyway. And all this was hashed out elsewhere long before the PCEHR program was ever suggested. It doesn't need to be raised again.

All the consultation paper you quoted says is "algorithms to derive the globally unique identifier documents need from non-globally unique identifiers are too unreliable to use in practice".

And all this time I spent writing this, I could've just read the consultation paper....

Bernard Robertson-Dunn said...

Grahame,

You and I are talking about different things. You are talking about the structure of identifiers, I'm talking about how they are created - the process: who does what, where it's done, when it's done, is it reasonable and acceptable?

As the Wikipedia reference you supplied says: "The identifier size and generation process need to be selected so as to make this (i.e. practical uniqueness) sufficiently improbable in practice".

Unless you can tell me how they propose to generate the identifiers, I suggest we drop this discussion before we bore and/or annoy everyone.

Grahame Grieve said...

@Bernard - both OIDs and UUIDs have associated procedures for their creation.

UUIDs - use the OS utilities or accepted libraries to generate a UUID.

OIDs - obtain a registered OID (usually via me), and then use your own internal serially generated primary key. OIDs are deterministic

As I said, the worst possible outcome for a non-unique document id identifier is that the PCEHR will reject the document. I admit, though, that I'm not entirely sure what will happen if the document setId is not unique, though I'm pretty sure that was on the system UAT.

I really think you're flogging a dead horse here. Generating identifiers must be done properly, the system enforces that, and is protected against failure to do it correctly. Implementers know how to generate unique numbers.

Bernard Robertson-Dunn said...

@Grahame,

"Generating identifiers must be done properly, the system enforces that, and is protected against failure to do it correctly."

Therein lies the potential problem.

The Pathology consultation paper says "It is proposed that the PCEHR Document ID is assigned at the same time that the report is generated"

AFAIK, the report will be generated outside the PCEHR, in a feeder system run by the pathology provider, and then it is uploaded into the PCEHR.

This appears to mean that the PCEHR Document ID is created outside the PCEHR. That means that the mechanisms you suggest, which would work very well in a single system, may very well not work across multiple systems. And I think I'm pretty safe in saying that health records and diagnostic reports in Australia are, and will be, held in many and multiple systems. This is why they have gone to a lot of trouble to create IHIs so that individuals can be identified uniquely across the multiple systems.

However, something seems wrong when it comes to uniquely identifying reports across multiple systems. It could be my ignorance or it could be the vagueness of the consultation paper.

It would appear that the consultation process relies heavily on previous involvement that was undertaken in 2009-2011. This makes it difficult for new people to become involved, especially when there are no references to other documents that provide the necessary context.

Even then I think that something like a "new field ‘PCEHR Document ID‘- not previously discussed during consultation" should have more details about how it would work.

Anonymous said...

Bernard, the PCEHR delivery team do and always have treated feeder systems as outside of scope and not their problem, this is the same with point to point. PCEHR is sucking the life out of eHealth. Now that Jane has left, it will be interesting if Paul Madden can hold the Jurisdiction to ransom or if they collectively take back control of NEHTA and eHealth, Paul Maddon comes across as being a bit of a wet fish So I know where my bet would be placed.

In defence of the consultation paper (which I have not read), I would say you and Graham have demonstrated its value in nurturing debate. Consultations are not about turning up with the answers and presenting 'look how clear we are', it is about carefully placing grey areas and slightly incomplete or incorrect subject matter, this stimulates debate and sets the platform for the stakeholders to come to some level of concession and leave feeling that the have been part of the solution, making change easier.

If you want to get into eHealth at a national level as you indicate in another post, this is something you will need to learn very quickly and skilfully. I think you would add value to the health debate, there are to many yes people and few left who challenge and provide thought through alternative solutions. Standards could have been that balance, however that too has been sadly polluted with self interests and self importance.

Anonymous said...

An interesting conversation, which illustrates just how hard it is for those of us who are not IT/ data boffins to understand and review the consultation papers.

Here are some questions I have:

I understand now that the PCEHR document ID is an OID and that it is therefore globally unique. The suggestion is that those systems that generate pathology reports must generate for each report a globally unique PCEHR document ID. But what if the report will never be posted to the PCEHR – e.g. because the patient does not have a PCEHR or does not want the report to be posted to the PCEHR? Is the proposal that the unique document id is generated just in case it is to be used in the PCEHR?

And presumably the pathology report can also be used in other e-health use cases, e.g. within a local shared electronic health record that manages chronic disease care across a range of care settings. What would be the ‘unique’ document identifier in this situation? Do we need to generate another unique identifier for every use of the document, or could we use the PCEHR document id – after all it is globally unique?

In ehealth in Austalia are we using these globally unique OIDs for other important e-health identifiers like IHIs and HPIs? I have an IHI but I reckon it is probably only unique within the context of Australian e-health. If it was taken out of context and plonked across the globe somewhere it might be someone’s bank account identifier or the phone number of a hairdressing salon or a winning lottery combination.

Oh it’s all so confusing, and if it can’t be explained clearly, and you need to be Graeme Grieve to understand it, then how can any of us have any confidence that we can adequately review the consultation paper? Or even that it will work?

Grahame Grieve said...

"I understand now that the PCEHR document ID is an OID and that it is therefore globally unique."

if only it were that easy :-(. It's actually an HL7 v3 II data type, which is a combination of a "root", which is either an OID or a UUID, and an "extension" which is optional. The combination of these must be globally unique. I'll have a go at explaining this (again) on my blog

"But what if the report will never be posted to the PCEHR – e.g. because the patient does not have a PCEHR or does not want the report to be posted to the PCEHR? Is the proposal that the unique document id is generated just in case it is to be used in the PCEHR?"

Probably. because it may become posted later - or information derived from it may be posted by a clinician later, even if the source is not. A single identifier will allow all these to be linked up - but here, I am speculating.

"What would be the ‘unique’ document identifier in this situation?"

most existing reports are identifed by an internal identifier that is only unique to the source lab, and there's no formal approach to give those identifiers a namespace so they can't clash; this puts the onus for namespacing them on the recipient, and they all do this differently.

"In ehealth in Austalia are we using these globally unique OIDs for other important e-health identifiers like IHIs and HPIs? I have an IHI but I reckon it is probably only unique within the context of Australian e-health."

An IHI is a 16 digit number that is only unique within the set of IHIs. When represented in a PCEHR clinical document, it's assigned the OID prefix "1.2.36.1.2001.1003.0." and that is globally unique.

"Oh it’s all so confusing, and if it can’t be explained clearly, and you need to be Graeme Grieve to understand it, then how can any of us have any confidence that we can adequately review the consultation paper? Or even that it will work?"

The complexity challenge is an ongoing one. Any clinical information system is complicated in multiple directions, but they still exist, and mostly work.

Eric Browne said...

Anonymous wrote "In ehealth in Austalia are we using these globally unique OIDs for other important e-health identifiers like IHIs and HPIs? I have an IHI but I reckon it is probably only unique within the context of Australian e-health."

These 16 digit individual and organisation identifiers are designed to work in an international context just like credit card numbers. See http://en.wikipedia.org/wiki/ISO/IEC_7812. Hence all Australian IHIs start with 800360.

Anonymous said...

What does OID stand for? Is it in the NEHTA documentation somewhere where I can't find it?

Grahame Grieve said...

for OIDs, see http://www.oid-info.com/

Anonymous said...

a search in Google (input=OID) returns a number of results, the first Eric's citation, the second Grahame's citation... how hard can it be?

Grahame Grieve said...

explanation here: www.healthintersections.com.au/?p=2160