Tuesday, July 26, 2016

Here Is Something The ADHA Should Start Working To Address As Soon As Possible. It Will Be A Long Job.

This report appeared a little while ago:

Reality check - reliable national data from general practice electronic health records

Deeble Institute Issues Brief No. 18
14 July 2016
Here is the link:
Here is the Executive Summary:
“Since 1998, data about general practice activity in Australia has been collected, analysed and disseminated through the Bettering the Evaluation and Care of Health (BEACH) program. BEACH has provided valuable information about how general practice has changed over time, the impact of policy on practice and general practitioner (GP) professional development, and is the most reliable national source of data on GP activity. However, its cross-sectional design precludes comparison of outcomes of different approaches to care.
It is estimated that 96% of GPs currently use computers for clinical purposes. However, some GPs only use Electronic Health Records (EHR) for part of their clinical work, such as prescribing or ordering pathology tests. Others are paperless and only use EHRs, but even in these circumstances the EHRs themselves lack the structure to reliably link management actions to a patient problem. There are at least eight EHRs used in general practice, each developed independently and structured differently.
In short, there are no nationally agreed and implemented standards for EHRs in Australia, in three areas:
·         EHR structure (including linkages)
·         data element names and definitions
·         use of clinical terminology and classifications.
Therefore it is not possible to reliably export standardised data from general practice EHRs of a sufficient quality to be used for clinical and research purposes.
With current policy focuses on data linkage, integration of care, improved use of the My Health Record (formerly the PCEHR) and attempts to use EHR data for research, the need for a reliable source of data from general practice EHRs has never been higher. Unfortunately there is no ‘quick fix’ solution, but the issues can be addressed with a targeted work program to address the three underlying problem areas.
This Issues Brief describes four steps required to produce high quality data from general practice EHRs:
1.      A defined EHR data model that links related data elements
2.      Consistent data element labels and definitions across EHRs
3.      Use of standardised clinical terminology and classifications
4.      Accreditation of GP electronic health records.
This recommended program of work requires a national, cohesive approach, involving stakeholders from government, professional organisations, the EHR software industry and organisations that use data from general practice.”
All of this becomes even more important given the GP Research Infrastructure has recently been defunded by the Commonwealth government, which is going to greatly degrade our knowledge of what is going on in General Practice (GP) over time.
The report comes from a research perspective so does not emphasise additional points I would make about ensuring the usability and robustness of compliant systems so they are well used by GP.
The obvious issue here is that GP Systems are provided and supported by (often small) vendors who may be hesitant to invest in work that leads to less differentiation and market distinction for their product.
We all agree that for both interoperability and research all these steps are needed. Just how this can be achieved will be a major challenge for ADHA!


Bernard Robertson-Dunn said...

Now there's a thing.

On July 24, 2016 2:56 PM I commented:

'Looking at the the supposed PCEHR "conceptual architecture", the one thing missing is "person" '

Today we get

"... the EHRs themselves lack the structure to reliably link management actions to a patient problem"


"This Issues Brief describes four steps required to produce high quality data from general practice EHRs:

1. A defined EHR data model that links related data elements"

Total agreement as to what is missing. - a consistent and comprehensive health data model of the very thing that is at the centre of health care - the patient.

In the case of eHRs, there is no consistency. In the case of the PCEHR/MyHR it doesn't even exist.

The first thing is for ADHA to even understand that there is a problem. The second is for them to come up with some sort of strategy to solve that problem.

The so called Digital Health Strategy that Dept of Health woefully tried to come up with and of which ADHA now has carriage, will make interesting reading.

It will answer, immediately, the question is eHealth going anywhere in this country? Or is it so stuck in the mud, no progress is possible, just more wasted money?

Andrew McIntyre said...

Thats why we created this profile
"Virtual Medical Record (vMR) for Clinical Decision Support – HL7 V2 Profile (V2IG_CDS_VMR_R1_I1_2010MAY) "

It allows a snapshot suitable for patient record transfer and research etc. Had we not been totally distracted by NEHTA, the PCEHR and the follies we have been occupied with something like this could be working now.

In reality we are in desperate need of eHealth to assess the long term outcome of people on various drugs, like statins etc and we should be collecting data suitable for research as part of our every day work, but we are spinning wheels playing with every trendy technology and shifting focus at the will of DOHA.

We have not progressed in over 10 years, which is why so many eHealth people subscribe to this blog, there is a huge level of frustration at lost opportunity, with government part of the problem rather than the solution.

john scott said...

Colleagues I sense we are now starting to shift our gaze from what is wrong to what elements could/ or should be the focus of our attention in regard to a shared understanding of a new way forward.

Bernard's point is well made. We cannot support integration and improved patient journeys unless and until we put the person receiving care at the centre of the Mission. We need to place the doctor and other clinicians of the care team at the centre of the System for reasons I outlined in an earlier contribution.

We can pick up on Andrew's point about a Virtual Medical Record as a concept to recognize the perspectives of different clinicians--most obvious with patients with chronic diseases. Whether we create actual shared medical records or we create effective health communication services to respond to requests for information about the person is then a matter of strategy and tactics.

The openEHR initiative and Archetypes provides a way to address clinical modelling and semantic integrity.

These steps and others require two steps to be taken:
1. Recognition that the way we are proceeding, particularly at the Commonwealth level, needs to change;
2. We need to help the Commonwealth Health Minister and her Department to make the case for change; this includes providing a more compelling narrative and transition strategy.

At some point those of us who see the need for change and see a way forward have to be prepared to stand with the Minister and even in front of the Minister when such a change opportunity comes into view. The issues here carry implications for our health system that are too profound for the necessary changes to fall victim to the media circus and blame game.

Peter said...

The first three of these could have come straight from the basic principles of Information Architecture. They are not unique to Health.
A common data model and a data dictionary are pre-requisites for meaningful integration between systems whatever the industry. Most industries have their own with significant overlap since customer relations are much the same everywhere. ARTS is the standard for Retail, TGOSS for telecommunications and finance obviously have a very strict format for all communications - including the associated privacy/security requirements.
For eHealth, there is a need to not only create a common data model but also, more difficult, get the different EHR vendors to comply. Internal processing can use whatever makes sense as long as the interfaces follow the standard. That, I thought, was the point of NEHTA (and now ADHA).
As I see it, a large part of the problem is that the Health industry seem to feel the need to re-invent the wheel - and to follow constraints imposed by paper records even when they no longer apply.

Dr Ian Colclough said...

July 27, 2016 9:10 AM --- "but also, more difficult, get the different EHR vendors to comply.
Which means a new way of thinking about how to perceive the problem is the first step. The second step involves a different way of defining the strategies required to get 'some' successes under the belt.

Bernard Robertson-Dunn said...

John, with you all the way.

Peter, I half agree. Where I disagree is that a single data model is required.

Health care is far more complex than retail, telecommunications or finance. Health care requires multiple models depending on the nature of the patient's problems and the complexities of the treatment (see below)

I suggest we need two things:

A few (no more than five) observations about the MyHR that demonstrate clearly to a non-technical audience what is wrong with the current approach

A few (no more than five) what I would call architecture decisions that a non-technical audience would understand.


The current system suffers from the following:

1. It is a simplistic automation of existing manual health record keeping. That means a) problems with such a system have been carried forward into the new environment. b) opportunities for better and more efficient health care have been missed and in some cases are now very difficult

2. It duplicates clinician’s existing systems, which opens up the significant possibilities of data inconsistency and/or incomplete data in the system

3. A deliberate, conscious and time consuming effort is required by clinicians in entering data into the system, ensuring data consistency and managing patient expectations. Furthermore, clinicians who enter the data are the least likely t gain any benefit from their efforts

4. Making the patient the focus of the system means that the needs of the real users (clinicians) are compromised. It also means that the patient can hide crucial data and misunderstand data in their health record.

5. The centralised database, direct patient access and patient control approach results in unnecessary risks to patient privacy.

Architectural decisions:

1. The real user is the clinician. Their needs should be the highest priority. They are users of health information, their role in creating, updating and managing data should be minimised

2. Patients should only access their data through the health clinician who is the prime user

3. Duplicated data should be avoided at all costs, thus a distributed system is required

4. A health information infrastructure comprising three major components is required:

a. Identifiers (individual and providers)

b. Secure, guaranteed and reliable messaging

c. A search function for identifying where data is located, but which retains no patient data

d. Access to data and its interchange is on a need-to-know basis

e. An information architecture/model that standardises data interchange and supports multiple health data models.

That last one is the key to innovation and progress.

IMHO progress in health care will only come from improved understanding of the processes involved in the human body and in delivering health care. These require multiple information models supported by data standards.

If you think of a relatively simple thing like a patient’s weight. There are a couple of easy measurements - mass, BMI. However, the causes of overweight and of achieving weight reduction are highly complex, involve multiple inter-relating factors and are impacted by many external factors. Data is necessary but not sufficient to understand causes and predict the efficacy of treatment

I predict that future health care will incorporate dynamic patient models as well as dynamic models of health care/treatment. Very few of these exist, certainly at the point of care.

These models will develop over time. A sophisticated health care ecosystem should accommodate models such as these but not mandate them and be independent of them.

These models should be implemented in the satellite systems that live next to clinicians and pathology service providers etc.

These are only suggestion, but we have to start somewhere.


Andrew McIntyre said...

We don't need any new technology to make it work, we just need to use the technology we have in place in the ways it was intended and to maximize its potential, which is vastly under used. Yes we need models, common models, but pathology has managed to transmit a vast array of atomic data in HL7 V2 and clinical data can be transmitted in similar ways. Its also backward compatible with existing implementations. Those implementations also need to comply with reasonable standards but the last thing we need is to advocate a move to another technology, that has been the problem all along. Maximizing the existing investment is the best way forward. Any new technology will most likely suffer the same fate as what we have unless we actually commit to proper implementations and if we are going to do that we should leverage the HL7V2 interoperability that already exists.

Peter said...

I more than half agree with you :-). I didn't mean to suggest that a single model was sufficient. None of examples listed have a single data model for all areas of the business - I think the logistics model has about a dozen different level 0 entities, hundreds of level 1 etc. Each line of business (c.f. medical discipline) has their own but there is a meta-model which links the key objects together if and when it is needed.
I do have a problem with the concept that Health care is not like everyone else. I have heard the same statement from Retail, Education, Defense, Finance, Telecommunications and Logistics. I do agree that Health is a complex area, but I think it would benefit from identifying and leveraging similarities rather than emphasising the differences.
As Andrew says - the necessary technology already exists, it is just that the Health industry is not taking advantage of it. To my mind there is a great number of authorities involved, but not a lot of experts.
Yes, yes and yes to your examples. As I have said before, MyHR should operate at a Master Data Management tool not a central repository. That is, it holds only a basic summary itself - along with the ability to reach into separate EHR, prescription, referral, pathology etc. etc. etc. databases to pull back only the information required, when required, and based on the role of the requester.
Retrieval would need to be authorised by all players: the data provider who holds the information, the clinician who will use it and, yes, the patient to which it belongs.
Note this does NOT mean the patient gets to see the records themselves. Tools exist, on the market now, for a person to grant access without being allowed into the records themselves. Basically it means two keys, held by different people (clinician and patient) are necessary to unlock the record.
Getting back to the data model - the key design is the common model for interfaces. A minimal set of data with extensions to allow for different content types within a standard format - e.g. BLOB (binary large object) sections for images etc. If a reasonable standard, i.e. not just a copy of paper forms, is defined then it makes sense for any new product to use the format internally as well and the standard takes over by default. Integration is the key here and a common message model is the key to inter-operability.

Bernard Robertson-Dunn said...

To be clear, I'm not talking about the business model of health care, I'm talking about models of the human body in order to make good measurements and then health care decision.

I'm most familiar with the electrical activity in the human intestine (my PhD).

Electrical activity comes from ion exchange between the inside and outside of smooth muscle cells. Those ions include sodium, chlorine and potassium. Salt (also sodium and chlorine) is present in blood and other areas of the body. Collectively, intestinal smooth muscle cells spontaneously oscillate. It was postulated that phase differences between the local areas of oscillation give rise to peristalsis - the contraction of intestinal muscles responsible for mixing and moving food down the intestine.

If a patient has a problem with digestion it could be caused by many factors, one of which might be ion exchanges across cell membrane, nerve problems (there are local nerve nexuses that impose some autonomic control of the intestine), too much or too little salt, imbalances of other chemicals, problems with bacterial gut biome etc etc.

A clinician needs to have a model of the normal functioning of these inter-relating, interacting systems so that they can acquire data (i.e. tests and scans), identify problems and recommend treatment. The treatment may have unwanted side effects so they need a model of the action of the recommended treatment in order to ensure that unwanted problems are not created, interactions with existing treatments are not created and that the patient can cope.

These are the models I'm referring to.

Health information systems need to support this type of model. At the moment Health Record systems mainly support GPs and their internal mental models.

IMHO, in the future, health care decision making needs to move away from the intuitive craftsman approach that was the only option in the past.

Medicine moved away from the guesswork prevalent before science provided an informed and evidence based approach. This also created indicators of "cause and effect". Health care now needs much more sophisticated models of cause and effect.

These will be developed over time to support clinicians. Flexibility and adaptability in information systems is required to support them.

MyHR is a newly developed relic of the past era.

We need models that have value to the clinician. (S)he is the one with the problem of how to help the patient. If we are not solving their problem we are wasting time, money and opportunities.

This (how to help the patient) is also the answer to David More's fundamental question "what problem are you trying to solve?".

Bernard Robertson-Dunn said...

Peter said:

"I do have a problem with the concept that Health care is not like everyone else."

Professor Jeffrey Braithwaite, founding director of the Australian Institute of Health Innovation at Macquarie University said:

"Modern healthcare systems are perhaps the most complex systems humanity has ever built."


Take your pick.

john scott said...

Colleagues I suggest we have at least four separate conversations at play in this conversation. To my mind these are:

1. Nature and role of electronic medical repositories in the context of an exploding knowledge base and technology capability and the resultant growth in sub-specialization;

2. Nature and role of electronic health records as they relate to informing and supporting the patient journey;

3. Nature and extent of reporting for policy and research purposes; this should be separate from billing so that the data is meaningful to clinicians and the policy reflects an assurance about the quality of the data;

4. Nature and purpose of interactions with patients and carers; the admin items are pretty straightforward, however the clinical issues are another thing altogether.

Providing a patient with the latest pathology results is akin to asking them to understand the meaning of various terms before they can begin to make sense of what they actually mean for their care. Our focus, I suggest, would be better targeted if we looked at what the test results MEAN in terms of the next steps--that is, next steps by the clinician and next steps by the patient. In Renal I have had a low level and background conversation underway for a couple of years about the value of augmenting path results with a new section which covers: a) what the clinician might suggest as next clinical steps; and b) what next step(s) the patient might take. The emphasis is on action.

When we step into the world of human communications we necessarily have to address language, cognitive ability, mode of delivery (e.g. dealing with the blind), etc.
This is healthcare's reality; it is only to a limited extent the reality of the commercial world. Duty of Care is a powerful discriminator.

Hence, I agree entirely with sentiments expressed by Jeffrey Braithwaite.

Bernard Robertson-Dunn said...

If ADHA is serious about outcomes rather than pursuing a private agenda, here's something worth looking at:

Patients in England can now view their GP record online

"Official figures reveal that more than 95% of GP practices are now set up to offer online access to detailed GP records including test results and diagnoses as well as referrals, immunisations, procedures and medications history, up from just 3% of practices in January this year."

I haven't looked at the system in detail, but as an approach I think it is well worth looking into.

It has low GP overhead, the GP is directly responsible for quality and accuracy, there's no national infrastructure or central database.

It doesn't do much for population health data, but trying to solve two problems at once can sometimes mean you don't solve either very well.

Bernard Robertson-Dunn said...

The UK approach also means that there is little or no need for ADHA, eHealth Legislation, Call Centres or GP training in the use of MyHR.

It could also be expanded into hospital systems giving a more comprehensive coverage.

The one use that MyHR might have - getting at a patient's health record in the case of emergencies could be addressed by expanding the IHI database to include the patient's GPs. That way emergency workers could have immediate access to a patient's full health record (GP + Hospital) using the same interfaces as the patient uses.

The IHI database could be created/updated automatically by GP/hospital systems.

Emergency services would then have an immediate indication of when a patient had last seen a health professional and who it was.

I may have missed something, but it seems a lot cheaper, better and more flexible approach than MyHR. But it does mean the government can;t get at everyone's health data. All they would know is that you had seen a health professional but not what for. Medicare and PBS details would be accessible but that's the case now.

Andrew McIntyre said...

One of the GPs I work with regularly email results to patients, when they ask for it. She says she gets rid of the red abnormals when they are insignificant as 5% of results will be abnormal normally. If she doesn't do this she gets lots of anxious patients worrying why there Bicarbonate is 1 below the normal range.

So she sends the results with some explanatory text and removes the "alarm" markers when they are of no consequence. I suspect this will be an issue if the PCEHR ever gets used....