This report on a speech in New York appeared a day or so ago.
Martin Bowles: using health information to improve care
by Martin Bowles
04.12.2015
Australia’s chief health bureaucrat is excited by the opportunities for technology in healthcare. My Health Record “can improve treatment decisions, reduce adverse events, avoidable hospital admissions and reduce duplication of medical tests”.
It is an exciting time to be a steward of the Australian health system. Our government has committed to reforms across the system: to improve co-ordination of care; efficiency of care; and sustainability of the system to improve health outcomes for all Australians.
I firmly believe data and information technology can provide breakthrough opportunities to enable and support this change. Adopting new treatment technology has never been a problem for health providers and professionals in Australia.
But changing their thinking about data and analytics and how this can help the patient and the system has proven to be more challenging.
Right now the Australian government has multiple reform processes under way which will encourage and support such a paradigm shift. When successful they will deliver the greatest improvements in our system in decades. We have reform activities in almost all parts of our system:
- Primary health care;
- Medical Benefits Schedule reviews;
- Mental health;
- Private health insurance;
- Pharmaceutical Benefits Schedule;
- Aged care;
- Digital health; and
- Jurisdictional responsibilities.
The results of these reviews will make our very good health system, which is under pressure, much better for patients’ health and also more sustainable. They will put data, analytics, evaluation and research at the centre of our policy thinking and our healthcare system.
A major challenge for achieving change in Australian health care is the complexity of the system. This can lead to duplication of services, gaps in care and poor integration of care. None of which are good for patients.
Adding to the complexity is the nation’s geographic vastness and low population density, which brings with it the challenge of meeting the needs of people living in rural and remote areas. Fragmentation and disconnect exists between:
- Primary, secondary, and aged care;
- Between private and public services;
- Across the different types of care that we might receive; and
- Between our states and the federal government.
Further challenges faced by the Australian health system are largely driven by:
- Rising consumer expectations;
- Costs rising faster than growth in our GDP;
- Rapid advances in health technologies with associated increases in costs; and
- An increasing need across the health system to deal with chronic and complex conditions.
The question is how to change it to make it more effective. We are looking at new structures and models for funding and delivering healthcare services. But we are also requiring improved collection, analysis of data and above all exchange of information across the system to drive better system performance and patient outcomes. This applies to individual patient information, and information that can be aggregated to provider, regional and national level.
Changes being made now will greatly improve our analysis of the information available at each level, advance our knowledge and understanding of the health of the nation, and facilitate new ways of delivering better healthcare, in ways never before envisaged.
Our Medicare system and Pharmaceutical Benefits Scheme provide excellent data on use of health care services. But we do not use it to its full potential. This is partly because of the complexity. But it is also because different parts of the system own and protect the Pharmaceutical Benefits Schedule, Medicare benefits, private, and public hospital data and have tended to block, rather than clear the way for, data sharing.
When I was first appointed I naively asked whether I could provide the Pharmaceutical Benefits Schedule and Medical Benefits Schedule data to our states and territories, I was given 100 pages of legal advice around privacy explaining why this was impossible.
When I then said — I want to do this. How could we make it happen? I received just four pages of advice. They now have the data. You need to ask the right question.
So it is possible. It requires the will, and the goodwill. Both of which now exist. It has not been seriously attempted before, but now we at the federal level are starting to share information with the other levels of government in Australia. We’ve still got a way to go.
There is a great deal more here:
Anyone who is interested in Australian e-Health needs to read this speech very closely.
Basically Mr Bowles is hoping that the use of data analytics on the information held in the various Commonwealth data-bases and the PCEHR will allow him to re-model the Australian Health System into a more efficient and less costly health system.
There is a germ of truth in the idea that if you could use information to remove or reduce the variability of clinical practice you could improve both the quality and effectiveness, as well as the cost of healthcare.
This is not a new idea by any means and has been worked at in the UK and US (and in Australia) for many years. (Under names such as evidence based medicine, clinical pathways etc. etc.)
The trick to doing this is have clinicians change their practice to do more of the things with their patients that makes a positive difference and do less of the things that either are ineffective or harmful.
How you do this requires a multipronged strategy - where analytics is just a way to identify and track the problem. One way of actually making a difference is to provide clinical decision support (CDS) at the point of clinical decision making - among a range of other professional development and educational initiatives.
Sadly support of clinical systems and CDS etc. seems to be off the agenda.
To me this is very, very stupid. A National E-Health Strategy needs to be a balanced framework that supports clinicians, software providers and technical staff as well as the central Department and its needs. If this is not the route taken then failure beckons and it is very unlikely the benefits sought will be achieved.
Shows what sort of direction you get when you have inwardly focussed bureaucrats who are disconnected from the real world, and who fail to consult and listen properly.
Pretty sad.
Do read the full article and let us all know - via comments what you think. To assist there is a short summary article found here:
http://www.itnews.com.au/news/how-the-govt-is-using-medicare-data-to-improve-the-health-system-412690
http://www.itnews.com.au/news/how-the-govt-is-using-medicare-data-to-improve-the-health-system-412690
David.
33 comments:
While I agree with your statements that there is a lack of emphasis here on CDS etc, I believe that we are a long way from being able to realise the benefits that will arise from those efforts. We have some (patchy) routinely collected 'health' data but it is only masquerading as 'clinically relevant data'. The data we have is not necessarily (or sufficiently) the data we need. Nonetheless, it could be made more broadly available (adequately de-identified etc) for some re-use. But sadly it is mainly 'counting data'. Mostly about encounters and services provided. How many of this, how much of that, when, who, where.... and not why? what was the Dx, Px, what was the level of complexity, of this clinical case versus that one, what's the likely differential Dx, what's the potential prognosis? what hasn't been done for this patient? missed? not followed up?.... I am not convinced that we have the data that can support these sorts of proper clinical data analytics, data mining, knowledge discovery. YET! It's possible of course, but let's see what the future holds.
I don't fully agree. Unless the systems used by clinicians add serious value to the quality and safety of what is going on in the provision of care the analytics will be little more than the activity counters you cite!
David.
It is not clear to me David whether you think Martin Bowles is making a lot of sense or not.
I think he is speaking unbalanced rubbish that ignores the quality of the sources of the data he hope to mine.
David.
I agree with everything David has said and most of the first comment - they are certainly not incompatible.
The data is questionable - all heath information is questionable - that's what health data is - hypothesis, opinion, point in time, context dependent. The Health Infomatics people who do not acknowledge this are living in cloud cuckoo land.
The main purpose of health information is to inform clinical decision making. Applications that can assist in this area are neither easy to create or overly common. Some do exist, but I doubt that they are very sophisticated. But that doesn't mean that innovative effort should not be directed that this area. Maybe the MSIA could promote such an effort.
If we look at the PCEHR, it is likely to do more harm than good. Too much bad information available to too many people who don't need it.
Martin Bowles is probably reading a script written for him by one of our favourite but misguided public servants. Neither understand the complexities of health information, health as system or the stupidities of the simplistic approaches that they are promoting.
The only way past a roadblock like the department of health is to go round it.
A glorious and rousing speech, the speechwriter should be proud, the way the unattainable has already been transferred to others as a source of failure is a wonder to behold. Not to worried at least we have another 10 years before anyone needs to pickup the pieces
I am impressed by
When I was first appointed I naively asked whether I could provide the Pharmaceutical Benefits Schedule and Medical Benefits Schedule data to our states and territories, I was given 100 pages of legal advice around privacy explaining why this was impossible.
When I then said — I want to do this. How could we make it happen? I received just four pages of advice. They now have the data. You need to ask the right question.
So it is possible. It requires the will, and the goodwill. Both of which now exist. It has not been seriously attempted before, ...
The person who was sitting at the helm of "not been seriously attempted before" is now at Finance. If Bowles is game to give Halton a bit of curry, he may well be useful in his role.
Anyway, after boring on about the need for a national identity management scheme for 30 years, I've paid out for Estonian e-Residency.
Is Estonia as good as The Republic of Elbonia?
Seriously though: "When I then said — I want to do this. How could we make it happen? I received just four pages of advice."
I wonder what those pages contained. Something like - break the law? Ignore the law, nobody is going to sue us? Or change the law such that Australians' privacy is deeply weakened?
One gets very suspicious with this lot.
Yes, David [December 06, 2015 4:26 PM]description 'unbalanced rubbish' is close to the mark. There is a lot of sweet motherhood which people have been writing now for over a decade. The more they write it, the more they believe it and the more they ignore the reality.
How to harness industry to solve the problem, develop and deploy a relevant, secure and reliable system, never gets a mention by Martin Bowles or any other bureaucrat, because they do not understand this vital piece of the puzzle.
December 06, 2015 6:19 PM said .... harness industry to solve the problem, develop and deploy a relevant, secure and reliable system, never gets a mention .......
There is a very simple reason which accounts for that/ Put simply, it is that the bureaucracy and the policy makers (well intentioned though they maybe) do not understand the industry and do not know how to lead this very complex industry. No-one should be surprised by that. Instead, they use the crudest of tools (the carrot and stick approach) in the misguided belief that will get the desired results.
I have a firm belief that governments role is to provide governance as they clearly cannot lead. To have useful decision support we need high quality atomic data. Given the transfer of care that happens all the time this means that the messages that transmit that data must be high quality and atomic. Currently we have atomic data in much, but not all pathology and the quality is patchy, but not bad considering the problems we have. High quality atomic pathology messages are not reliably handled or displayed correctly in many cases so what we have it data that is hacked to ensure the doctor who views it is happy, which often means its display orientated.
What the government has done is pay large PMS vendors money to implement the governments pet projects ie PCEHR related material which has meant they have no time and no financial incentive to actually produce standards compliant implementations for the stuff that is actually useful. (ie the Standards we already have) We have been hijacked. Mind you they only paid the large vendors, I am not sure that is fair to smaller vendors who could have better solutions in areas, but miss out on funding from government. Mind you its careful what you wish for, as I am not sure its an advantage.
In the clinical arena we seem to have virtually no atomic data by most vendors (In messages, which equals transfer of care). Most clinical messages are blobs of rtf, which is not very interoperable in fact. (MS word made sure of this, so everyone needed word). There has been no focus on improving the quality of clinical documents to a level that supports decision support.
Now you can do a lot of decision support with v2 pathology, but its planned to be a pdf in the PCEHR, which is just criminal. So the PCEHR pathology is pretty much useless for any automated quality activity.
In other areas the government just says comply with the Australian Standards or you cannot market or sell your product and the standards are mostly consensus standards. This would work in eHealth, despite the fact that the consultants would disagree as they would have much less work. Its not sexy and it would take time and would increase the cost of software. Really they should have given most of the Billion Dollars they spent to the Medical and related Professions on the proviso they only used tested compliant software, which would have significantly increased in price. Its clear the software needs to be better and that costs significant dollars and its paid for by the providers so they could have directed the money to compliant products. This is much fairer than giving the 4 largest PMS companies a bundle of dollars to develop the pet projects of NEHTA/DOHA.
However until the software is better and the data compliant and atomic we are almost by definition going nowhere. Its a matter of getting out of the ground. No one likes spending big money on foundations, but experience tells you not to build on a swamp, and thats whats been done. It would take Yoda to lift the PCEHR out of the swamp, and unless December 17 brings surprises he is already dead.
David re this comment: [Australia’s chief health bureaucrat is excited by the opportunities for technology in healthcare. My Health Record “can improve treatment decisions, reduce adverse events, avoidable hospital admissions and reduce duplication of medical tests”].
I am just back from 3 months in Chile so I thought I would make a comment.
The whole concept of information management as the cornerstone for effective health care reform is not new.
The following is taken from an article by Bill Tierney and others.
Although health care is considered a service profession, most of what clinicians do is manage information. They collect data (take a history,perform a physical examination, read reports, look up laboratory data, read x-rays), record data (write visit notes, operative reports, prescriptions, and diagnostic test results), transmit data (via
telephone, paper or electronic charts, and email), process information to arrive at a likely diagnosis (or hierarchy of possible diagnoses), and initiate treatment. This initial chain of information management is then followed by
additional cycles of data collection, management, and processing to monitor and adjust care. Thus, information is not a necessary adjunct to care, it is care, and effective patient management requires effective management of patients’ clinical data. According to Gonzalo Vecina Neto, head of the Brazilian National Health Regulatory
Agency, “There is no health without management, and there is no management without information.”(1)
Translated into to real-life health care delivery and management in developing and developed economies. (2) More at www.openmrs.org
1. Tierney WM., Kanter AS., Fraser HSF., C. B. A Toolkit For eHealth Partnerships in Low-Income Nations. Health Affairs. 2010;29(2):268-73.
2. Braitstein P, Einterz RM, Sidle JE, Kimaiyo S, Tierney W. "Talkin' about a revolution": How electronic health records can facilitate the scale-up of HIV care and treatment and catalyze primary care in resource-constrained settings. J Acquir Immune Defic Syndr. 2009;52 Suppl 1:S54-7. Epub 2009/11/05.
3. 3. Hannan T, Solving a Health Information Management Problem. An International Success Story. World Hospitals and Health Services. Vol. 51. No. 2. 2015 pp43-37.
@Andrew:
"This would work in eHealth, despite the fact that the consultants would disagree as they would have much less work"
2 comments:
* I think that you're wrong to describe this as a 'software' problem. Though there are certainly purely software components to this - e.g. bad escaping and legacy design - there's a wider problem in that the standards have not been bedded down by extensive testing, and (particularly now) are not responsive to lessons learned during the implementation phase. Further, additional clinical agreements and discipline are needed to use the data elements as described. This is not just a software issue. And there's a legacy data issue too
* So I don't think that 'the consultants' would have less work with this approach. Different work, maybe.
I agree about pathology, btw, but the outcome comes from extensive consultation with the industry - exactly what people are always asking for. Be careful what you ask for...
Martin Bowles suggests that many don't see healthcare as a system'.
He goes on to suggest that we need to:
1. shift from hoarding to sharing;
2. work out the right incentives when chronic disease is commonplace;
3. manage the interfaces better; and,
4. achieve a health system that is sustainable.
He states: "we have to have combined goodwill and courage to actually take the decisions that we need to take. And, to look for innovative ways of doing things into the future so we don't leave a lesser system behind."
I take this as a door opener to a different and complementary conversation.
I suggest we could beneficially canvas a separate and to some extent parallel path.
This would operate under the umbrella of our mutual interest in a sustainable health system.
It would take a whole-of-health system perspective; not just the public system.
It would accept that digitization of health represents a strategic opportunity.
It would not assume that the knowledge and assumptions that underpin our current digitization strategies are correct.
It would seek out new knowledge and new instruments for productivity.
These new instruments need to enable our health system to be better connected and integrated.
The benefits of proceeding in this manner are:
1. we admit that there are innovations that could make a difference;
2. we accept that change needs to be organic; and,
3. we accept that we need a more informed view of what operating in a digital environment means for a connected and integrated health system.
We need to shift from organisation-centric to network-centric thinking and acting.
Re comments from Andrew and Terry about the importance and place of information.
1. Information is critical and needed to make decisions - however, it's the decisions that matter most.
2. Health information is fuzzy at best and misleading at worst. Just about everything in a health record (apart from the date and nature of an incident) can change. Test results can be wrong, especially if taken out of context; diagnoses can be wrong and/or incomplete; prescribed medication is not the same as what the patient actually took; allergies can come and go; even gender can change, etc, etc.
Ian's warning "Be careful what you ask for..." is very true. In the case of the PCEHR, our bureaucrats thought that the medical profession was asking for more data - a health record system. What the medical profession should have asked for is help in better managing the flood of fuzzy data they often get when trying to make decisions.
What did they get? A system that is just going to make things worse.
Hi Graham,
we spend most of our support time working around issues where applications can't handle compliant data or can't produce it. Often is basic things like word wrapping, document identifiers or escaping that causes issues but its always at a base level. I know we could have much richer content but the fact that there are bugs in base levels means going to the next level is not a real option. Its using data for decision support and improving workflow that I want to enable and this is limited in the real world by issues of poor implementation/compliance currently.
I am not sure just who "they" consulted about pathology, but it was obviously someone who only cared about what the data looked like when displayed. We are so far off the track that people have lost track of the whole point of doing eHealth. PDF healthcare is not that different to PIT based healthcare, just a whole lot more resource hungry. The point is that if we had governance of standards compliance then they would have been confident of what the data looked like when displayed by any application. Loss of atomic data is a fatal flaw. There are a lot of fatal wounds, but rather than operate on the patient to fix the internal damage they have employed cosmetic surgery to try and hide the entry wound.
I was disappointed that pathology data - the one set of data that we could actually get to consensus about, and really usefully exchange now - was excluded from the PCEHR. I'll leave it to the direct participants to explain why we ended up where we did on that question.
If anyone is interested in the root cause of this disaster:
http://www.smh.com.au/business/the-economy/broken-public-service-leads-to-broken-governance-20151205-glg8y1.html
"The move from Yes, Minister to Be It On Your Own Head, Minister has come in stages, starting with the decision of the Whitlam government to allow ministers a much greater personal staff of (unaccountable) policy advisers and media managers. The Fraser government perpetuated this "reform" with enthusiasm.
The Hawke-Keating government's main contribution was to replace "permanent heads" of departments with department secretaries on five-year contracts. After five years heading one department you'd be moved to heading another.
Thirty-odd years of this and now senior bureaucrats rarely stay long in any department, but climb the ladder by moving from department to department.
They've gone from being long-experienced experts in particular policy areas to "universal managers". I may not know much about health or finance, but I know how to run a department. Great."
> I am not sure just who "they" consulted about pathology,
> but it was obviously someone who only cared about what
> the data looked like when displayed.
At a guess - pathologists? Or is that too obvious an answer?
This is an edited speech delivered by Martin Bowles to the New York E-Health Collaborative Forum at the Academy of Sciences on December 3.
Find out who else was there, and who took Bowles out to lunch. May be a clue to the new direction and who may be invited to collaborate.
I am sure it was pathologists who are very picky about the display of their reports. However pathologists are not the users of pathology and their concerns are real and should have been solved years ago by making sure that the display of pathology was compliant with the standard. This is the alternative solution. There are multiple interests to listen to, the users of pathology probably want it to display well also, but pdf is the worst solution to the problem and negates any future decision support. You don't solve display non compliance by trashing the data.
Last time I check I have > 4 million reports dating back to the first PIT results that were produced in Queensland. If every one on those was pdf the storage requirements would be enormous. I can also do cumulative results on atomic data back to the start of HL7 V2 around 2000. What are you going to be able to do at reports in the PCEHR in 2030 - look at them!
We need to store data in a format that is useful to future decision support, that was done in the year 2000. That was before the government stepped in to help.
Hi Trevor
And the answer is.... CSC!
http://www.csc.com/health_services/success_stories/31450-csc_supports_new_york_ehealth_collaborative_in_trial_implementation
Karen,
I'd be very careful quoting from a vendor's own website. Especially a vendor with this sort of track record:
http://www.techweekeurope.co.uk/workspace/nhs-2-2bn-csc-npfit-failed-contract-118941?PageSpeed=noscript
Accenture also have a track dubious track record when it comes to claiming success.
According to their website (written before December 2014) they helped the NSW Department of Education to a triumph:
https://www.accenture.com/lv-en/success-new-south-wales-department-education-communities-learning.aspx
Unfortunately, the NSW Audit office didn't agree:
"The Department of Education and Communities (the Department) commenced the learning Management and Business Reform (LMBR) program in 2006. It was expected to be completed by December 2014 at a total cost of $483 million.
...
Executive summary
Audit conclusion
Some parts of the three major components of the LMBR program have been delivered to
different areas of the Department. However, complete implementation is behind schedule and over budget, and the Department has yet to demonstrate that it will achieve the expected benefits."
http://www.audit.nsw.gov.au/ArticleDocuments349/01_LMBR_Full_Report.pdf.aspx
16 December 2014
Not commenting on the success or otherwise of any CSC projects....
The NY ehealth thing is indeed a collaborative venture, but the event Bowles addressed appears to have been a private function, and there is no information available about it on its website
http://www.nyehealth.org/
CSC is providing the support for its trial implementation
Here, CSC has sat on the federal e-health sidelines since the commencement of the PCEHR juggernaut.
But it was a pretty keen contender previously and of course holds huge e-health business worldwide.
Perhaps Bowles took a look, while he was there
BTW, Departmental luminaries are supposed to post a record of their official speeches here...
http://www.dpmc.gov.au/sites/default/files/publications/aust_govt_public_data_policy_statement_1.pdf
This part of the website a pretty dry and dusty corner
I really don't know how Health is going to cope with Open Data!
>I really don't know how Health is going to cope with Open Data! (Karen D.)
I'd like to know how Health copes with its own data. I heard from a reliable source, within the last 5 years, that the call desk had to log in, separately, to several systems to access data.
I remain amused that the standard "authentication" question for medicos using the prescription shopping service is the medical school attended by the inquirer.
Trevor
re: "I'd like to know how Health copes with its own data"
The ANAO isn't impressed with health's records management performance:
http://anao.gov.au/Publications/Audit-Reports/2015-2016/Records-Management-in-Health
From that ANAO report (Thanks Bernard)
1.18 The TRIM EDRMS is identified in Health’s internal corporate policies as the official ‘recordkeeping system’ capable of storing departmental records. As at 30 April 2015 Health had more than 205 other separate computer systems to assist staff in carrying out their duties to deliver its 10 business outcomes. Health has not assessed all of these systems to identify if they contain potential records.
Trevor
This was the opinion of the Privacy Foundation:
http://www.privacy.org.au/Media/MR-HlthData-151202.rtf
"Something's not quite right. It's manifesting itself as a lack of trust in the department and a major risk to the privacy of all Australians. Their claim to be trustworthy custodians of intimate health information in the PCEHR looks less and less credible. For a government skeptical of Big Government, the push for this struggling bureaucracy to take control of private medical records from clinicians and patients is surprising and has never been properly justified."
Not just "lack of trust", Bernard.
Seems to me if interoperability was the key requirement of NEHR when it was being conceived in the 80s, the landscape would be completely different. Instead, the needs of big public hospitals for costing (hence, billing) data drove all. New systems were installed without having workable (and provable) interoperability. With that paradigm as key driver, it's no wonder the systems are even more entrenched into impregnable silos. Brigham and Women’s Hospital switch to Epic may be a good example, if ever the truth is allowed to escape.
Another example. In the 90s a few public hospital pathology departments enrolled in a costing comparison program run by College of American Pathologists, aimed at getting data on 'billable tests'. One of them was able to show the inputs were about break-even with outputs - excluding pathologists salaries. No need for a sociologist to predict how that went.
So, a Health Department with over 200 separate computer systems was charged with making the NEHR. The System Operator who presided over that impossibility is now at Finance, leaving Martin Bowles on a career-ending trajectory. From Public service looks to slash $4b corporate bill.
Finance will lead an effort to revolutionise the bureaucracy's corporate functions in an attempt to slash the public service's $4 billion a year administration bill.
Trevor,
IMHO, interoperability is a technical enabler; it is not a health problem in its own right.
The real and only problem is better health decision making.
If that is where you start then, you discover that what is needed is relevant, complete (to the decision being made) and reliable health information. A health record is often not relevant, complete or reliable. The PCEHR contains too much data that is irrelevant, incomplete, unreliable and accessible by too many people who shouldn't be able to see it.
The PCEHR started in the wrong place (a dumb, partial health record) and went downhill from there.
One has to feel sorry for Martin Bowles. He has inherited the poison chalice. It is clear he just doesn't get 'IT'.
His utterings reflect the same old bureaucratic rhetoric over and over again. It's not Martin's fault though; he hasn't been in the job long enough to have to accept responsibility for what his predecessor left behind her.
He says "I firmly believe data and information technology can provide breakthrough opportunities to enable and support this change. Adopting new treatment technology has never been a problem for health providers and professionals in Australia. But changing their thinking about data and analytics and how this can help the patient and the system has proven to be more challenging."
Mixing and matching buzzwords like "changing their thinking about data and analytics" shows how very little he understands. The problem is - he needs help and doesn't know what help he needs so he turns to those in the bureaucracy who have created this eHealth mess nationally in the misguided belief they will be able to help him. One has to feel deeply sorry for him.
I don't feel at all sorry for Martin Bowles. The first thing he should have done was some due-diligence on the Department. To rely only on the advice from bureaucrats in the department is sheer folly. It's a sign of an administrator, not a leader.
He needs a view from outside that he can compare with that coming from those with a vested interest - the insiders.
Post a Comment