National Health IT Board - Draft National Health IT Plan

There's some discussion on Computer World about the draft plan that might be of interest  - around reducing the amount of data collected by the MoH and including patient information from private organisations:
 
http://computerworld.co.nz/news.nsf/news/draft-plan-addresses-health-it-failures

Hi Graeme
Nice to see the plan up on Hive, and I hope we get some more discussion.
 
From the plan we have the phases set out, and in each phase some lists of what needs to be done.
 
Lets bring some pressure to bear on ther hospitals on the transfer of health information between organisations.  There are curerntly 3 off the shelf products from vendors that can be bought andd plugged in for e-refferals. Lcoally it is very hard to get traction from the DHB on this.  These solutions all offer security and a process audit so we, the referrer, knows what has happeended to the information.  I find it intolerable that we are getting faxed notifications that a referral has been prioritised 4 months after it has been sent in.  Clearly this is not just IT, but process as well.  Given the primary secondary interface is by definition dealing with people in a state of flux with problems that may be unstable, the ongoing risk that ensues from current processes is in my view unethical.
Also, the lumping of primary/community together is a misnomer.  General practice prt of primary has, on the whole, a high level of IT/IM use and infrastructure.  The community providers have little e capability evident, adn they are often part of the secondary services.  need to help that part of the sector get up to speed.
 
Elsewhere in the document is the stressing of patient centered information.  I've no problem with this, as you know I'm an early adopter of the only current NZ patient record.  From a practical perspective however, (and not very PC to point this out), the tools the health worker uses every day 8-10 hours a day need to be focussed on the main user.  By all means have a patient held record, but hopefully the PMS vendors will continue to keep an eye on their main customer!  Clearly I can see great things with a patient held EHR (eg they can manage their own recalls with the data uploaded from our systems > less nurse time > nurse free to do more clinically relevant work), but the reality is that most of the high cost high needs patients are not likely to be accessing the record a great deal, (perhaps their families and caregivers will - I hope so).

Phase 1 Consolidate, Co-operate and Foundation

(Jan 2010 – December 2011)
Focuses on delivering key solutions to provide the:

easy access to health information

transfer of health information between healthcare organisations

capture of clinical event information into a regional clinical data repository

improvement of primary healthcare practice management systems

consolidation of systems used in secondary and tertiary settings into regional platforms.

Phase 2 Shared Care

 
(March 2010 – December 2014):
Will commence with a design and proof of concept phase and will deliver a Shared Care capability covering:

Patient Vitals: This is historical patient information eg. Demographics, problems, medications, warnings and access to more detailed e-events (labs, radiology and medication history)

Care plans: This is future patient information that captures the plan for the patient’s future course of care. The single plan incorporates the need for a multi-disciplinary approach to support integrated care

Decision support: Supporting reference information to support

Although only skimmed this so far, what I saw as very disappointing was that there is not reference to a population tool particularly for primary care.  A tool that enables for example the following

• identify patients who they should be seeing and invite them to attend,  
• to monitor how they are doing in terms of achieving targets eg. diabetes get checked 
• to present the data in a simple effective way to the all the team(s)  
• to report the results in an aggregated way to PHOs/DHB 

Will enable the health sector to continually monitor how we re actually doing against targets.  It also enables analyzes at different levels so that it is useful information at the coal face right through to the National picture.For much of the BSMC targets this is a building block to enable these to be robust measures and useful to support change.  Gwalarn
 

I agree with this point. If we could identify patients pro-actively, their health could be monitored in a better way, i.e. help them before the illness gets the better of them. In the example given, patients' chronic illnesses could be monitored by employers, who can encourage the visits to appointments, as this will increase productivity at the workplace, and also portray the employer as a caring employer. Unfortunately, the employer is often discarded from being an important contributor to monitoring health. The irony is that most patients spend more time with work collegues, who are aware of the condition of the person, and who could assist in the monitoring thereof. Of course it is a sensitive issue, but so is absenteeism, and even worse, presenteeism.Dieter (www.equeue.co.nz)

I had first quick read on proposal and wanted to read again to go through in details. before that thought to share few concerns in my mind. 

1. Having two phases is good idea. however looks second phase is more challenging one referred as shared care. I would suggest to go for initial design period for both phases and then break the implementation into two phases.
2. Implementation of any new software to clinic may results issues as happened to many international EHR implementations. During my post graduate, what i proposed was doing part by part introducing system. for example initially data capture system at discharge/any place where it not break the main care activity. And then move to other areas. 
3. As already many clinics/hospitals using some form of software to manage their activities. It will be good to come up with platform with this and deliver set of webservices/api to integrate with any software, which open up integration path for many existing applications use this new platform. And important thing is EHR model selection too.
4. Have seen many place in the proposal the word centralised.. Need to be sure that system distributed and managed across the country without any single point failure as this human life business and can not tolerate on any downtime :-)

Hi Folks,
 
We were developing two separate discussions so I've moved the comments from the forum over to here so all the comments are in one place.
 
Please continue to post in this thread with your comments on document.
 
Thanks,
 
Chris

Hi Graeme
 
I fully support the direction the National Health IT Board is taking and the requirement to have a National IT Architecture Framework supported by a virtual sector architect team and centers of excellence teams. The thinking of not reinventing the wheel by using best examples currently available, following industry best practice and standards is also very important.
 
I would like to make a few comments regarding the National Health IT Plan.
 

• The plan is structured as an Enterprise Architecture Vision and is a statement of architecture work. This follows the work pattern that has been carried out with HealthBase.
• Section 4 Our Approach. HealthBase uses Transformation Plans to allow an evolutionary approach, supported with the use of Standardized Technology; this reduces complexity by using a minimum component set. The Transformational Plans allow us to move from the current state to a Target State satisfying the Health Sector strategic objectives.
• Section 5 Governance. I would have expected to see an Architecture Governance Group in the “Hold Accountable” area of the diagram, such as the HealthBase Governance Group.
• Section 6.3 Conceptual Model. The New State Conceptual Model represents the Architectures found in HealthBase these are aligned with TOGAF. (Business Architecture, Application Architecture, Data Architecture and Technology Architecture). We have made sure all HealthBase work is aligned with TOGAF, but is constrained to support the NZ Health Sector specific needs. HealthBase captures these Architectures in Reference Architecture Documents; these documents hold the Architecture Building Blocks.
• The remainder of section 6 focuses mainly on projects to deliver clinical needs. There is no holistic plan on how these projects align to satisfy the target state, the target state lacks a conceptual model to provide sufficient depth of understanding. We need to clearly articulate how this transformation will be delivered by the 28 projects noted in the plan; otherwise there is a danger of a fragmented Health System.

 
I would be pleased to show you through HealthBase Enterprise Architecture and how this could help the National Health IT Board.
 
Regards
 
Alan Le Maitre

Thanks for your comments.  i am reading these with interest.

An interesting read. I would suggest that there is no logical difference between the data in the messages (e.g. GP2GP) and the clinical data repository. It will be tempting to ignore this and get some aspects moving quickly but the expense of having many different models and implementations going on  at the same time is huge (and usually fails).
Clinician friendly is good - but the reality is that benefits and savings are more likely if it is clinician driven. This is not such a winner politically but I believe it is the truth. If there is not strong clinician buy in then it is not a health record. All consolidated data has to match the same logical conceptual representations or it will be meaningless so it has to come from the health record at some point.
I will be in NZ discussing openEHR and comprehensive, clinician led approaches to health information next week and hope to learn more of what is going on in your (always) innovative part of the world.
Sam Heard

 
Hi Graeme
It was interesting to read the draft plan and in general I believe it is a good start.  I would like to make two points that I think will help New Zealand avoid some of the mistakes that have been made in other jurisdicitons, including Australia.

• One of my main concerns is that the plan confuses the EHR with an application and puts too much emphasis on applications.  It suggests that picking a winner in the application space is problematic, because 'the script is only partly written'.  My take on this would be always to put the emphasis back on the information.  Applications will always be out of date, old technology, proprietary etc etc.  The EHR is THE INFORMATION not the application.  If NZ gets the information right - open, non proprietary, and future proof, then the applications will look after themselves and vendors can all compete on a level playing field for look and feel as well as functionality.  This document is too skewed towards the applications as the driver for the information and not the other way around.  If this is not understood, then NZ is likely to make the same mistakes that UK and other jurisdictions such as the states in Australia have made - buy an application from a vendor to solve the problem.  This leads to vendor lock in and poor interoperability.  Messaging has consistently been shown to NOT solve this problem.
• Clinicians need to be front and centre here.  If the clinical information is not owned and governed by clinicians, then the data is likely to be a lot less useful and will be skewed towards the needs of Govt. data collections rather than patient care.  If we get this right, then the data for national collections will be a natural by product of the collection of clinical data during patient care.

I hope this is helpful.
regards Hugh Leslie, Clinical Director, Ocean Informatics

Further to some of the previous comments from primary care, there doesnt seem to be much mention of population health at all in the document. Interestingly, the American Health IT Strategic Plan (which can be seen here http://healthit.hhs.gov/portal/server.pt?open=512&objID=1239&parentname=... ) has two overarching goals - patient focused health care and population health - and all objectives relate to both of those goals. I would also be interested to see the evaluation of the plan's implementation with respect to population health outcomes.
Robyn Whittaker, Clinical Trials Research Unit, University of Auckland

Can I say at the outset the National Health IT Board is very aware of the importance of usibility of health information systems. Optimising the time it takes clinicians to access, add or update information is top of mind as the board reviews priorities and the design of health IT solutions. An interoperability approach to health information means that there will always be a number of information systems that need to work together. These systems must work together seemlessly, to maintain the effectiveness of the 'systems' for clinicians.
I agree that there are three vendors in the market providing e-referrals solutions. We have seen a significant improvement in uptake since HISAC took up the promotion of e-referrals to DHBs in February 2009. The next major implementation is in the Northern Region where referrals for 25 specialties will be in place by the end of the calendar year.
Experience in the Hutt Valley, Canterbury and Northland shows that referrals work best where there is a trusted relationship between Primary and Secondary clinicians. I can only assume that your DHB has not yet progressed an e-referrals project. The NHITB will be working through the DHB CE Information Group to keep the pressure on DHBs to deliver on e-referrals (and e-discharges) by the end of 2011. The Board will be openly and publically reporting progress against its plan (including the implementation of e-referrals) by DHB.
Now, I have a different view re. Primary and Community healthcare organisations from an information point of view. I accept that General Practice is well ahead in the use of information solutions, however, the style of healthcare IT solutions required by community based organisations are more similar to primary care systems, than secondary. This is because both Primary and Community healthcare organisations are "network" based, requiring relationships with individuals (and their whanau/families). In contrast Secondary care organisations are generally healthcare service providers, reacting to the clients who are referred to them, or who access their acute services.
Graeme
 

I am pleased that there have been a couple of comments about the importance and value of population health information. Pleased be assured the board have not missed this area. We are approaching this subject under the focus area 'Quality Information for Primary Healthcare'.
It is a broad area including: outcome reporting, quality improvment activity, decision support, and proactive pubic health, such as screening programmes. In the first instance we are working with the Royal NZ College of General Practitioners to review current programmes such as the Quality and Outcomes Framework (QOF) programme that was introduced in the UK in 2004 and the Primary healthcare Performance Programme (PPP) we have running here in NZ.
The goal is to gain agreement by clinicians on the core primary healthcare information that in the first instance is required to be collected to support the delivery of healthcare services. When aggregated at a local, regional or national level, this information can enable performance improvement in clinical practice (through peer review etc.) and support population based preventative activity.
The overall goal is for health information to be available for people to develop and test insights that drive an ongoing understanding of cause and effect in terms of improving healthcare outcomes.
I will look to make the subject of 'population health' more clear in the next version of the plan.

I am pleased that you have recognised an Enterprise Architecture framework within the plan. 
To be successful with the plan, we need to have an Architecture Group that reflects broad understanding of the business, information, software and technology models that are in use by healthcare organisations in NZ. As well as the ability to develop a desired target state that is made up of a number of reference models.
Tony Cooke, CIO for Hutt Valley DHB and seconded to the NHITB team, has been engaging with a group of architects in the sector through Darren Douglas, which I understand is the same group who are behind Health Base.
While we are happy with progress my understanding is that this group is still in a formulative stage and has been focusing on key development projects such as the Recepient and Provider Identity (Replacement of NHI/HOI etc.).
I am happy to look at formalising this group in the near future.

Over the next 2 years we will support 3 shared care demonstration projects that will develop health records (both a historic view and a future view).
The projects will be selected by finding subjects where clinicians strongly agree there is a need for multi-disciplined clinical access to health information, that will improve the patients journey, experience and health outcomes. My philosophy is based on thinking about Health Records in the same way we think about the development of web sites, they need to be sticky to ensure people, in this case clinicians, come back often and continue to gain real value on each visit.
Sam, I look forward to having a chance to talk when you are in NZ next week, especially any experience/information you can provide on clinicians leading such projects.
 

Hugh, you make a good comment about the plans over emphasis on applications, the risk being that the absolute importance of information is lost.
The plan is trying to strike a balance between the current way health IT is discussed and understood in the sector in NZ (which is very application orientated) with a future health IT model based on models of care and clinical pathways, supported by coherent and consistent health information.
The other balancing item is a recoginition that the NZ market is not large when compared internationally and will struggle to attract new market entrants. The National Health IT Board has a goal to create an open, transparent health IT market that is guided by priorities as set out in the plan.
You second point about clinical governance of health information is heard loud and clear. NZ has plateaued in terms of national collections of clinical information, because the information has been collected as a by product of claims and payment processes.
I fully agree with your comment "...then the data for national collections will be a natural by product of the collection of clinical data during patient care."
We want to define our information requirements in terms of clinical value and then aggregate it at local, regional and national levels to meet different user needs. Information Governance with clinical and consumer leadership requires significant development to make sure it is in place to support these new clinical information models in the future.
Very helpful, thanks

I followed the link to the American Health IT Strategy Plan - it is a really good strategy. However, It reminded me that we did not want to repeat the 'strategy' discussion again in the National Health IT Plan. The main problem in NZ is not strategy, we have had good health IT strategies, we just don't implement them!
Population health is included in the plan under the focus area - Quality Information for Primary Healthcare. Have a look at my earlier response (don't worry I only posted it earlier tonight). We will make this area much stronger in the final plan.
Cheers,
 

Hear hear Graeme re implementation.
There are still things from WAVE that we need to implement.  If we had actually done that over the past 10 years the stagnation you have mentioned elsewhere would not exist.  We pretty much know what needs to be done.  There are a multitude of work items that need to be done with a view to underlying standards.  I sincerely hope that in 12 months we will see significant progress.

Hi Graeme
 
The group of architects you refer to was originally the HealthBase Architecture Review Board and I agree this group has made good progress and is being driven by Tony and Darren. It is good to hear you are looking at formalizing this group.
 
My main concern is the Architectural approach and content in the National Health IT Plan. I find it is project driven and not taking a Health Sector holistic view, there is an architectural layer missing between the business strategic objectives and the projects delivering the plan. Historically the Health Sector has been project driven and this has caused the fragmentation, causing the problems we are currently faced with. The National Health IT Plan needs to show an architectural approach that is based on industry best practice and provides an architectural transformation plan to allow an agreed path to the target state. This needs to be supported by a standards information base and reference architectures that further define the target state.  I don’t feel this architectural approach is described in the National Health IT Plan and this concerns me as I feel it will limit the plans chance of success.
 
 
Regards
 
 
Alan Le Maitre

Alan, The use of Focus Areas and care pathways is an attempt to make the link between models of care and information solutions.We have purposely not gone into a great deal of architecture planning as we wanted this document to have wide readership.
if you want to carry this conversation going online can you be more specific and share some examples. If not, then it would be good to receive feedback directly through the architecture team via Tony and Darren.
Thanks,
 

This is a great initiative but the document seems to miss a key goal in looking at how IT could reduce demand on the NZ health system – rather than just looking at how to deal with the demand. Certainly, I agree it needs to be clinically led but this does not necessarily equate to clinician led – ideally we want to keep at risk people from needing to be treated by clinicians. I think the plan needs to recognise that there are fragmented and vested interests in the delivery side of health care in this country and look to the Health Ministry for evidence based input on heath priorities - something they do fairly well. At present Public Health and Population Health areas in New Zealand are not supported by a coherent Information Strategy and yet are highly dependant on good information to target efforts. Have these areas been canvassed on their needs? My related concern is the plan makes statements about what is needed but there is little evidence to give confidence that these are the things that should be the priorities for step improvements in the sector. There is a lot of good research and experience in this area – has it been used in development of the plan. All said, this is still a long overdue initiative and I think all involved in Health Informatics should support its success. R W Crawford

I have been reading the plan and these comments with interest and believe that this provides a very good vision as a starting point.  However, as a clinician now working in IT, I have some concerns that there is insufficient stress on the fact that for better or worse, health services are here to provide health care to patients/clients/service users.  Health IT should be working to ensure that the information  is collected, stored and available to clinicians in such a way as to assist the clinician to provide the best possible care to users of health services.   Having worked as a DHB clinician providing community based services, I would like to see some focus on mobility of information.  Ability to carry my files with me and update them at the point of care rather than having to come back to the office to do this.  Laptops are all very well and good but there are other options out there now and there use should be part of this strategy.Finally I have to say it is great to be able to be part of the process of change.  To be able to give feedback and feel listened to is critical if people are to embrace the changes that are going to happen over the next few years.  It is not about being the one to say what should and should not happen, it is about having the opportunity to "have a say".  It is critical that part of the implementation of this plan is good change management.

Certainly I agree that using information for quality care delivery should be an important focus of the Plan. What I am saying is that it cannot be the only focus. I also agree that clinicians must have a say in things that impact how care is delivered. The issue I have with the current Plan document is that it does not cover the wider uses of Health information to inform health policy and support population based programs such as immunization and screening. Every western democracy has to face issues, not of how to better deliver care, but of how to deliver care at all given projected demand. New Zealand will not be able to build enough hospitals or recruit enough nurses unless we find ways to keep people out of hospital as they age. This is the challenge the government faces when every DHB is saying they need more money and is struggling to maintain, let alone enhance services. Maybe the Plan is not supposed to address this – a question for Mr Osborne – but I believe that unless the Plan considers the relative priority of investment, as the pressure comes on it will be very hard to drive the improvements proposed. However, it may just be about introducing a common health record. Perhaps that’s really all we need?

Thanks for your comment, we have had feedback that the plan would benefit from a greater focus on Public Health ad Population Health. Please don't assume we have access to all the research in this area. We are always happy to receive links and suggestions re. research and evidence.
Re. Public Health, we are pretty well informed as one of the members of our team was involved in the creation of the current system used by the Hutt Valley DHB Public Health team. We are likely to require the three Public Health teams to work together to create a national approach to defining information to support Public Health activity.
Population Health on the other hand is covered in some ways in the draft plan in the areas of screening and some of the well-child checks/immunisation. The key area we want to improve is access to information for clinicians to work proactiviely with their populations. Currently the Ministry is doing some good work to make national information available for PHOs and primary health organisations by working with third party health information companies - Dr Info and Health Stats. In the future we want information available locally, regionally and nationally.
 
 

Good comment, the plan is currently silent on the need for mobile access to information. Don't take this as a negative, we understand the importance of mobile access to clinicians and patients. We will make a comment about this in the final plan.
Now, a story to tell, last week I meet a paediatrician clinican in Gisborne. He was showing off a tablet computer that he was able to access the hospital based clincial record over a wireless network. In fact he was also able to access labs and radiology results.As he explained instead of taking parents back to his office or crowding around the PC at the nurse station, he could have a conversation in the corridor or by the bed.
The IT team have found a way of linking the clinical workstation with the PACS system. This means that the Dr can show parents on a flat screen a clinical record and a full motion echocardiogram - really impressive. Add to this that the tablet can have a mobile phone SIM card in the back and a clinician can take the access on the road outside the hospital - now we are talking.
It takes a small, innovative team in Gisborne to solve this challenge - I was really impressed, and I know the parents are getting a better experience.

The NIHR has a lot of good stuff (http://www.nihr.ac.uk/Pages/default.aspx). There is not a lot of New Zealand specific research but here are a few links to resources and papers I have found credible and useful in understanding the NZ context for health IT:-
http://www.moh.govt.nz/dataandstatistics
http://www.treasury.govt.nz/publications/research-policy/wp/2006/06-05/twp06-05.pdf
http://www.hinz.org.nz/journal/2009/12/Investigating-the-Potential-to-Improve-the-Free-Flow-of-Health-Care-Information-----a-Survey-of-New-Zealand-Clinicians-and-Health-Providers/1012
http://www.nzma.org.nz/journal/117-1198/977/
http://www.moh.govt.nz/library.html
I am sure others can point to additional sources.

 
The National Health IT Plan of the ITHealthBoard is comprehensive, realistic and ambitious. However the current draft does not address the single most important critical issue and risk: the standard to be chosen as basis for an EHR-architecture that meets the desired outcomes for exchange, sharing and storing of information. The standard selected will influence the national EHR-architecture substantially. Without knowledge about the standard to be used the feasibility of the National Health IT Plan cannot be judged.
 
When the Sector Architecture makes an incorrect choice regarding the standards to be used, the integration and deployment will need a substantial part of the projected resources for innovation and replacement of existing systems. The same goes for the extension of the current portfolio with IT-systems with functions critical for executing the National Health IT Plan. Projected timelines won’t be met.
 
The diagram on page 9 of the draft National Health IT Plan reflects the need for coordination. The standard to be used for exchange, sharing and storing of information will influence the aspects that need coordination and the intensiveness of the coordination. The standards to be used should on the one side make coordination more easy. On the other hand the standards should contribute to a transparent process for selecting the right future proof systems by decision-makers during the process of consolidation. Systems not capable of using the standard or systems not capable of creating added value out of the use of the standard probably won’t be future proof. 
 
It is our strong belief that innovation of the provision of healthcare is not driven by coordination, but is driven by a bottom-up approach facilitated by a dynamic IT-environment in which creative destruction can do its work: improvement of productivity by survival of the best fit-for-purpose IT-systems and ditto vendors. Standards are used in the European Community as a regulatory instrument to create a level playing field that boosts competitiveness by innovation.
 
This triggers the question: Why not use a suitable CEN/ISO standard to create a level playing field and clarity for all stakeholders involved?
 
The suitable standard we are referring to is the EN13606. This standard makes it possible to limit the number of interfaces for exchange, sharing and storing of information to the number of systems that need to be connected with regional Clinical Data Repositories (CDR). E.g. the same interface layer in a GP system, used for the connection with a regional CDR, is used for implementing:

•  GP2GP exchange;
•  continuum of care;
•  transfer of care for cross-enterprise shared care;
•  integration of clinical decision support using all information from different systems;
•  creating additions to current systems for patient centered treatment of chronic diseases.

 
Through the use of the EN13606 standard health information data is stored in a normalized way in the CDR and used via a middleware component implemented by vendors in their IT-systems or during the period of migration added on to these legacy systems. This middleware component presents to all IT-systems a flexible, adaptable, unified standardised EHR for shared care or Patient Summary.
The middleware assures the uniform and standardised uploading of data to the CDR and downloading of data from the CDR for use in the IT-systems. The fact that all IT-systems need to integrate this middleware is a small investment in comparison with implementing standardised messages for specific use-cases. Using the EN13606 it takes day’s/weeks per system to implement the middleware, depending on the information needs. In contrast it takes several years to achieve the same functionality with less semantic interoperability using the messages paradigm. E.g. implementing the epSOS patient summary for cross border exchange in Europe took less than one week. Time to deployment to achieve early gains is crucial for the sustained support by stakeholders for the National Health IT Plan. 
So what are the consequences of using different standards (approaches) in real life?
 
Present day IT-architectures are based on working (legacy) IT-systems creating archipelago's of data that cannot easily be exchanged. Thereby creating a situation where such facilitating IT-solutions do not support the co-operating healthcare provider. (hcp)
Presently used proprietary IT-systems rely on the message paradigm (Edifact or HL7 message standards as defined artifacts). These technical artifacts, that need to be implemented uniformly by the IT-industry, define rigid datasets in a technical IT-way. The uniform implementation of messages has been proven inflexible and time and money intensive. IT needs the process used by the Integrating Healthcare Enterprise (IHE) organisation taking at least 18 months for the preparation of  a profile and testing it.
 
Healthcare needs innovation supported by IT-solutions. Rigid solutions as provided by present day proprietary data silo’s and message standards have been proven not to support innovation at local scale at reasonable cost. When these systems are tailored to local needs the integration gets tailor-made too. However this shouldn’t be the reason for a national standard for all work processes. Best practices and evidence based medicine must be the driver. Since these drivers change rapidly over time due to new knowledge one needs as prime focus the requirement for the frequent implementation of updates to local standardised work processes.
 
New technology (based on an open International CEN/ISO standard EN13606) can create the situation that hcp’s play a pivotal critical role and have them involved in the change processes. Tools are available to help hcp’s define their shared care information needs and configure applications that support the clinical and business model behind it. The IT-industry does what it is good at: the processing of bits and bytes. An EN13606 based CDR creates a complete separation of concerns between the healthcare and IT-sector. Whatever the hcp’s define the IT-systems based on EN13606 can store, retrieve, archive and act upon.
 
Our knowledge about standards and lessons learned because of experiences due to actual use of the EN13606 at several places around the world and our knowledge about costs, implementation lead time and results of current approaches for national EHR-infrastructures using messages, developed at the turn of the century, are used for this high level response to the draft National Health IT Plan.
 
Summary:
Standards are mentioned in the draft document.
There is no clear statement that stresses the need for a national infrastructure for seamless semantic interoperability based on EN13606 based technology with a strict separation of concerns for hcp’s on one hand and the technical IT-industry on the other hand.
 
Using an International CEN/ISO standard has of consequence that parts of the whole architecture for implementing the National Health IT Plan already exists. They are built by different technology groups all around the world. This global usage of compatible components is a prerequisite to get to a level of commodity pricing for EHR services.
 
The approach of using the EN13606 supports in the best thinkable and comprehensive way the envisaged engagement with different stakeholders in the draft National Health IT Plan. 
 
ERS is happy to elaborate on the approach of using the EN13606 for implementing the National Health IT Plan with decision-makers, influencers and the directly affected persons and organizations. 
 
Gerard Freriks, director Electronic Record Services (ERS). Former chairman CEN/tc251 wg1
René Schippers, director Electronic Record Services (ERS)
 
g.freriks@e-recordservices.eu
 

Hi Gerard
 
I have read your response with great interest. I agree with your statements regarding the national infrastructure and standards and agree this architecture approach is needed. The HealthBase enterprise architecture group has put considerable effort into these areas. There is currently a conceptual model that uses data services with a service bus approach; this has been influenced by Infoway. The standards that are being considered are HL7 v3 and CDA, as they are current NZ standards, we have also considered openEHR. This is very similar to the approach you are proposing using EN13606. IHE is another consideration when we have agreed the standards.
 
Regards
 
Alan Le Maitre

Dear Alain,
A few words of caution. It is not correct to equate message based solutions with more object based solutions.
An EHR-Architecture based on messages (and HL7 v2, v3, CDA are) need a resource intensive production, implementation and testing phase. IHE is an good organisation that provides these services. 
Applications based on EN13606 have a major advantage. You do not need IHE, because experiences have shown that it is possible to implement a Patient Summary between two university hospitals in Spain in a matter of days.
The reason: Software that does the information gathering is generated on the spot using an Archetype Editor, a mapping tool that produces executable artifacts.
Gerard

Hi Gerard
 
Yes I do understand your words of caution and reason; NZ have considered both approaches and feel the message based approach is the more appropriate. The message based approach we are taking is a federated Model, with regional EHR repositories. It would be good to discuss this further, I will contact you directly.
 
Thanks
 
Alan Le Maitre

 
One last comment on the IT Plan. Most New Zealanders are fair minded and I suspect many are too polite to mention it but the content is let down by being poorly written. Normally I would not comment on this aspect but the document needs to be professionally credible and the standard of English is below that of document of this importance. Please get the next draft third party reviewed and edited – it deservers better.

Hi Graeme, others
 
I must say the most striking point for me is the phrase "core set of personal health information...". Unlike the previous strategy I think this is a realistic target for NZ given the high level of Health IT literacy - especially the clinicians. But there is hardly any explanation in the plan how to achieve this other than the mention about meta-data repository and clinical data repositories. So I'd say this aspect of the plan needs serious work. And keep in mind that 'the devil is in the EHR' and no matter how small or big, a core EHR is still an EHR.
 
My background is mostly development of clinical information systems, and I have been involved with standards around EHR. As someone having knowledge and experience on HL7 and openEHR/13606 (more on the latter) my one clear message would be not to get obsessed with standards at the moment. There is currently a lot of fuss and hearsay but hardly any evidence. On the other hand I believe lots of valuable evidence has already started to accumulate worldwide and in NZ too as projects such as GP2GP, eReferrals and the like start kicking off. I think the following two years or so will be quite intense. Perhaps daydreaming but in the meantime why not ask for a proof-of-concept shared care record based on the current requirements we have - say in 6-12 months or so? I suspect with a fraction of the cost for a national rollout we can learn a lot.
 
In the recent paper (Atalag K, Paton C, Kingsford D, Warren J. Putting Health Record Interoperability Standards to Work. electronic Journal of Health Informatics 2010;5(1)) we have looked at adoption patterns in different jurisdictions (warning already old information) and the bottom line was that there is no clear winner and that selection must be based on requirements and business cases. Inevitably the landscape will be a 'medley' and the challenge will be how to make them work together.
 
I have two recommendations to make:
1) Invest on getting the 'information' right: no matter which applications are to be selected this will always help and never get wasted. Many comments suggest clinician engagement as a critical factor. Yes this IS how you engage the clinicians - let them specify, design and govern 'their' health information. This shall form the foundation of  requirements elicitation and development for interoperability.  There is now a straightforward way to do modelling using Archetypes without making any serious commitment to openEHR/13606 or HL7. These models can be transformed from and to another. - although far from perfect at the moment due to limitations of HL7 v3 RIM.  This is the path many national programs are taking/considering. The proposed meta-data / concepts registry approach in the Plan will be quite sufficient for projects involving administrative and high level clinical concepts. However in phase two for establishing a proper and scalable shared care record it is unlikely to succeed if recent advances in this area are not carefully studied.  I can assure you that governing clinical knowledge is far more difficult than we can possibly envisage at the moment. Therefore a more advanced shared concept repository, capable of supporting clinical Archetypes built upon SNOMED and other nationally approved terminology will be a necessity. This simply implies that either terminology or clinical models won't be enough - they should be handled together. This is our best shot for achieving a high level of semantic interoperability. Indeed looking at the recent collaboration between IHTSDO and openEHR (IHTSDO and openEHR Begin Collaborative Work Programme) I wouldn't be surprised to see both SNOMED and Archetypes under a single umbrella soon. So insert a placeholder for such a repository when building the data dictionary and meta-data repository which some of the current projects mandate/need.
2) Fill the gap between best-of-breed (i.e. integrated vendor systems) and fit-for-purpose(clinical/ancillary systems, especially Access databases or other clinician led applications).
Healthcare profesionals will never buy in the former if they don’t clearly see they are getting the latter functionalities at ease. In practice this is all about the scary “clinical forms” which most big vendors are freaked out because they are so different, complicated and change all the time! But we need to get that 'valuable data' into a clinical data repository which can ensure that they are of high quality, accountable, safe and queryable. So how is this going to happen? Well only if we can provide the clinicians with tools which they can define, design and deploy own small databases and registries which will answer their myriad of research questions and provided all that convenience, workarounds etc. Is this possible? Yes - but a cautious one at the moment. The precursors are proposed as: 1) common terminology, 2) common information model, 3) common clinical models. Although the approach I recommended above does not require to make a committment to HL7/13606/openEHR, this one does because the information model is standard specific. I am not aware if and how that happens in HL7 space but with openEHR the clinicians and small vendors can workout their fit-for-purpose systems which can be interoperable or easily integrated with big systems - without any compromise on the interoperability or semantic coherence across regions/NZ. Well in theory of course - but there is at least a sound theory ;)
 
One other dimension is that we have to be realistic about what we can expect from vendors, especially in a tiny market like this. No vendor likes standards - unless it is beneficial or mandatory. The latter does not really apply to big vendors because standards need to fit what they can deliver anyways. Looking at the big vendors we really don't have much of a problem as they have to achieve and maintain a high level of quality in order to stay in the global market. However it is the small vendors or inhouse developments that we must make sure they deliver on the requirements. Enforcing standards, by means of testing and certification, can be an effective means to validate product quality and integration/interoperability capability.

Hello Graeme.  Interested in your comment that "the main problem in NZ is not strategy."  I'm aware (but not in detail) that there has been various health IT strategies hatched for the publicly-funded side of the healthcare sector.  Yes, that is a sizeable chunk of all healthcare but by no means all.  As I understand it, more elective surgery is delivered in private hospitals than in public ones.
The concept we're talking here about an electronic health record capturing "a core set of personal health information....regardless of the setting."  To provide a sufficiently comprehensive and reliable record for clinicians to work from, does that record not need to cover all of the individual's health issues, regardless of where they receive that treatment?
I know from personal experience that I have bounced from GP, to private specialist, to public hospital, and back to private specialist in relation to one particular health condition.  And the referral pathways become even more entangled when treatment for separate but potentially related health conditions are overlaid.
Is there a health IT strategy (or a health strategy for that matter) that embraces the private health sector?  In what ways does this draft National Health IT plan encompass healthcare providers outside the public funding streams (essentially DHB facilities/programmes and primary care)?  I can only see a single reference in the document to private hospitals and no reference to medical specialists in private practice.
The economists say we have to be smarter with our money, because we're not rich enough as a country to keep up with the rest of developed world when it comes to healthcare.  So let's be smart - let's make sure we have strategies and plans that focus on providing Kiwis with the best possible outcomes for the money they spend on health - not just their taxes but ACC levies, health insurance premiums and out-of-pocket expenses as well.
James Norman

Hi James,
Good question, before I answer I need to confess that I have held CIO roles at both Southern Cross Healthcare and ACC in my career. I mention this only so you know that I fully understand your question abut health information created in a private or publicly funded care setting.
As you stated, there is no difference from a health information point of view. I can put your mind at ease that our goal is to collect health information from both sources. Today, the regional laboratory results solution in Auckland (called TestSafe) captures lab results from private and public labs. We will continue the lead already created by this solution.
Graeme
 
 

Hi Graeme
I am a Kiwi clinician in Boston accompanying my wife Penny Andrew who is the NZ Harkness fellow 09/10. I have just completed the Harvard certificate program 'Leadership for Information Technology in Health Care' led by John Halamka and others so I would like to offer my two cents worth. Also just joined the finalists for NZ Focus on Health who were in Boston on their NZTE tour of USA - what a great lineup of Kiwi ingenuity!
Please note these comments are my personal views only and not necessarily the views of my clinical colleagues or HBDHB where I am employed. 
I commend you and your team on the plan. I think it would be good to be a bit more upbeat about the opportunity that exists for us in NZ to make a significant leap forward in HIT. We should recognise the convergence of 3 key factors that will allow this to happen:
1. IT technologies themselves
2. Smart informatics for proactive population and personal clinical decision intelligence
3. Clinical and public desire for improvement
I would like to see the title reflect this: "Enabling an integrated, personalised healthcare model"
The new state conceptual model looks good ( I like stacks) but one thing I don't see is the
incorporation of a learning system into the model. This is fundamental to how we will continuously improve
and I think it needs to be included or at least acknowledged in your plan.
Finally, I know we have plenty to do as clinicians to drive a shift in our thinking to deliver "better, sooner and more convenient care" but I believe there is now enough groundswell support for this within our hospitals and in primary care to make a difference. We  should however be held more accountable for leading this change than we 
currently are. Likewise we should take more responsibility for assisting your Board in implementing the HIT plan.
I'm looking forward to getting back and getting involved with this exciting work.
Kind regards
David

Hi Graeme - thanks for the opportunity to feedback on the plan.  I work as informatics analyst and integrator within a public health service, so perhaps bring a slightly different perspective from clinical and IT vendor/architecting streams.  From that perspective I would strongly support attention to what a previous poster described as "smart informatics for proactive ... intelligence".
 

As you would be aware, with the overall exception of infectious disease management, public health service business does not generally involve personal-line-level access to clinical data.  However, access to aggregate outputs from such data is at premium for public health surveillance, intelligence, analysis, and planning. In the field of clinical information, once again with the major exception of notifiable/communicable disease, PHS's are therefore generally in the position of consumers rather than contributors to clinical datasets, and I suppose my comments here need to be taken with that in mind.  Any strategy which potentially leads to improvements in access to on-demand, secure, integrated population-level datasets on health service inputs (in primary and secondary sectors) outcomes and determinants of health is very welcome from this point of view.  I believe this is at a premium, particularly with movement towards more distributed models of health service delivery which, whatever their merits in other respects, have the potential to silo and black-box service-level information. 
 
As a general point, I would like to support Hugh Leslie’s cautionary points above about planning tending towards application-specific emphasis and the perils of “(buying) an application from a vendor to solve the problem“ – an approach which has affected information systems implementations in the population/public health service field from time to time.  Hugh has also noted the problems caused in other jurisdictions by the big-bang application-led approach.  Open, non-proprietary standards or models for information provision and consumption are definitely the way forward here.  Other posters have reinforced this point by talking about interoperability and API / Web services approaches.  An engineering-led, single-solution-path paradigm for system design and implementation can lead down a blind alley in an environment where responsiveness and active management of processes and relationships can be more important.  I believe this point was reinforced during last year’s H1N1 incident where information, analysis and reporting flows performed better in those areas where there were strong, pre-existing, plans and relationships between different sectors and providers – IT infrastructure being only a important conditioning factor, but not an arbiter, of overall success.
 
As mentioned above, with the partial exception of communicable disease, much of the information space occupied by Public Health Services lies outside the scope of patient-centric electronic health records and is occupied instead by such concerns as environmental health (for example drinking water quality, biosecurity and quarantine, food safety, environmental risk management, hazardous substances), regulatory functions associated with alcohol and tobacco, and health promotion project work where the main desired outcome is prevention of chronic disease (nutrition and physical activity, smoking cessation for example).  In addition there is an increasing emphasis upon analysis, planning and interventions based upon intelligence from broader determinants of health such as service provision, inequalities, ethnicity and culture, transport and the built environment.
 
Clearly this population / public health information space lives quite outside the very proper scope and emphasis of the plan upon an integrated patient-centric information system for the delivery of personal health services.  In addition the scale of information systems and problems within population and public is much smaller and less time-constrained than those presented by personal health information.  However, in case they are of some interest, I will conclude with a couple of comments about the population / public health information space.
 
Historically, Public Health Services have been particularly impacted by nationally implemented systems which have been tightly focussed around the information needs and processes of a single business line, for example (but not restricted to), food safety, communicable disease, drinking water quality, smoking cessation, some screening programmes.  In each case these systems succeed to one degree or another from the vertical focus of that particular business line, but do not provide strong integrated informatics capability at a public health service level. Each system functions as a spike driven down through the operational space of the public health service and have historically lacked APIs or other simple capabilities around data exchange which allow for integration at the point of service delivery.  In some cases public health practitioners (public health specialists such as Medical Officers of Health, Health Protection Officers, health promoters, analysts, and others) will be required to input and extract information from a number of these disparate national systems.  Processes, service and contractual requirements for each one of these systems naturally tend to emphasise their specific requirements, impacting upon the informatics needs of the service as a whole.
 
On a conceptual level population and public health informatics issues are quite similar in some respects to those in personal health records.  In particular, systems are more likely to succeed where they provide adequate space to be practitioner and service provision led.  Additional factors in success are the ability for otherwise disparate systems to exchange data at appropriate levels of detail (where exchange is an operational requirement and not just a “nice-to-have”) and a degree of configurability / expandability, without vendor lock-in, which allows for innovation as well as inevitable changes in organisational structures, service provision processes and reporting and analysis requirements.
 
Once again, thanks for the opportunity to comment.
Chris Ambrose.
 

Thanks Chris for your response. I understand the public health perspective but what I was getting at was the trend that is occurring here in the US to 'marry' population health data with an individual patient's problem (patients like me) to better inform both the patient and provider on diagnosis and care plans. It's a natural progression of predictive modelling and is being used with good effect at Geisinger. 
Cheers
David
 

Thanks David, and apologies for taking your point out of context ...  neatly illustrates differences in  languages between personal and population-level health, and as noted I'm quite aware of tiptoe-ing across a field where population health issues are largely out-of-scope. On that point I was talking about turning the microscope upside down, as it were, using aggregate data products from information systems to generate public health intelligence, which in turn may inform planning, perhaps preventive strategies, eventually feeding back down to personal health status. 
Cheers !
Chris.

Hi Graeme,
It is great to have an opportunity to make some comments on this plan.  There have been some really good recommendations made so far.   As a clinician working in New Zealand and student within the field of health informatics, I wanted to offer a few humble suggestions to hopefully echo some of the important comments highlighted earlier and their relation to the proposed plan.
 

1.       Obtaining data for the National clinical data repository

The proposed plan looks towards implementing something that is achievable and deliver a sort of ‘shared care record’ through capturing information from discharge summaries and referrals.  The potential pit fall of this is at present this information only captures information at entry and exit to the hospital system.  
 
There is no mention of a plan to capture information for the national clinical data repository during an admission as a patient transfers from one team to another during their admission.  This places a large burden on the trainee interns and junior house surgeons that have to complete many of these detailed discharge summaries each week, who sometimes have to summarise two to four months worth of notes for a patient they have only looked after for a week or two.
 
At present the national standard for the content and format of these discharge summaries and communications is under development and should be finalised next month in June.  It is important to ensure these standards are enforced across the board from organization to organization so the information within these discharges contribute meaningfully to the national clinical data repository. 

 
 As part of the roll out there needs to be standardised education of junior doctors and trainee interns to ensure that the format of summaries are acceptable and standardised across all organisations.   This will need to be monitored and audited frequently to maintain high standards.  It may be useful to make use of the Physician Documentation Quality Instrument (PDQI) that was developed by Stetson et al (2008) to help monitor the quality of discharges(http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2442259/ )
 
  The issues around developing these records was highlighted nicely in the following link (http://tastethecloud.com/content/patient-health-records-reality-check) 

 2.       Managing the data within the National Clinical Data Repository

To develop an appropriate shared care record for patients, that is assessable. There needs to be more of emphasis on the development of ‘intelligent agents’ (IA) to help manage this information.  So that the information used to do a discharge letter or referral can be pre populated directly from the national clinical data repository or the hospitals EHR saving time.  Also by using these IA’s it could help facilitate future m-health and potential cloud computing options in a few years time.  Thereby increasing consumer access to their information.  In other words developing a system similar the Sundhed portal developed in Denmark as highlighted on page 11 of the draft plan (http://www.themobilehealthcrowd.com/?q=node/287) 
 

3.       Implementation issues

This is one of the most important issues in the success or failure of large scale IT programs.   There is a considerable body of evidence available to highlight how the socio technical aspects and change management approach chosen can have a significant impact on programs success.  
Pagliari (2005) highlighted some valuable lessons learnt from the Scottish experience of Electronic Clinical Communications Implementation Programme which had components very similar to phase one of this National health IT plan.(http://www.haps.bham.ac.uk/publichealth/cfhep/info/Pagliari_NPfIT.pdf ).

The draft document mentions that “The approach to implementation will be to trial innovations in different settings where they are most likely to succeed”.  This approach sounds like it will lead to a classic diffusion of innovation based approach to implementation that may not be as successful as alternative methods.    This was highlighted by Greenhalgh et al (2008) who highlighted the importance of shifting the change model from "technology push" to "socio-technical development" may improve the chances of successful implementation of programs.  (http://www.bmj.com/cgi/reprint/337/oct23_1/a1786).
The classic diffusion of innovation theory focuses mainly on factors that affect the speed of uptake of an innovation in terms of organizational and technological readiness. The main problem with this is it leads to more of a project management style of implementation and it fails to acknowledge the complexity and diversity of health facilities.  Callen et al (2008) put forward the contextual implementation model (CIM) that may be a better implementation model.  CIM focuses on contextual readiness of the organization, the unit/G.P practice and the individual.  The main difference is in the acknowledgement of the diversity of organizations, units and individuals which would in turn lead to more of a program management/socio technical approach.  They highlighted this well with a diagram that can be seen in their article. Available here: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2274802/pdf/255.S106750270700343X.main.pdf  

  4.       Mobile Health

It is great to hear of the entrepreneurial spirit of some clinicians and Graeme’s example of the paediatricians in Gisborne with iPad like applications highlights the growing trend towards the use of mobile devices to deliver health care.  There was no comment regarding the use of mobile applications within the current plan and we can only hope that the final plan will incorporate them and facilitate their use.The benefits of this were highlighted in a recent systematic review by Prgomet et al (2009) in JAMIA (available here http://jamia.bmj.com/content/16/6/792.full.html)

 Conclusion

 As a clinician and health informatic student it is really great to be able to contribute towards the discussion about the future plan for health IT within New Zealand and hopefully this wide spread engagement will continue through to the implementation stages.  Thank you for providing the opportunity to contribute and I look forward to seeing the final version.
 
Nathan Billing

Health Informatic Student and Senior Dietitian WDHB
Nathanbilling@gmail.com

 References 

Callen, J.L., Braithwaite, J., Westbrook, J.I., 2008.  Contextual implementation model a framework for assisting clinical information systems implementations.  Journal of the American Medical Informatics Association 15(2). Pp. 255-262. [Online] Available at: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2274802/pdf/255.S106750270700343X.main.pdf [Accessed 01/03/10] 
Greenhalgh, T., Stramer, K., Bratan, T., Byrne, E., Mohammad, Y., Wood, G., Russell, J, 2008.  Introduction of shared electronic records: multi-site case study using diffusion of innovation theory.  British Medical Journal 337:a1786 [online] Available at: http://www.bmj.com/cgi/content/full/337/oct23_1/a1786?q=rss_home [Assessed 24/05/10] 
Pagliari, C., 2005. Implementing the National Programme for IT: what can we learn from the Scottish  experience? Informatics in Primary Care 13:105–11. [Online]Available at: http://www.haps.bham.ac.uk/publichealth/cfhep/info/Pagliari_NPfIT.pdf [Assessed 25/05/10] 
Prgomet, M., Georgiou, A., Westbrook, J.I., 2009.  The Impact of Mobile Handheld Technology on Hospital Physicians' Work Practices and Patient Care: A Systematic Review.  Journal of the American Medical Informatics Association 16: pp792-801

I have not had time to review all the posts so this may of been touched on already.
I work in the aged care sector for a private health care provider. 
There is a large, rather wrinkly, elephant in the corner of the room that no one wants to address. We mention "aging populations" as a key global strategic driver (pg 4 of the plan) and we all know this is a significant issue but what are we doing about it?
The plan is a good draft but it appears to be missing some key stakeholder groups, NGO's. I am obviously focused on aged care but there are also other groups that are missing. I would be happy to provide more input but it appears the core "selected" stakeholders are government based, to the exclusion of NGO's. I know this is not the true case as articulated by Graeme last Wednesday (laying out the game plan) so we need to make it easier to have NGO's contribute.
This bias is well articulated on pg5 with the Ministers six health targets, from an aged perspective, what about Dementia? The whole person centered approach is valid through out the care continuum and our lives, but especially so where care becomes more continuous as we age. 
Getting a core set of electronic health records will require the "peripheral" NGO / Private providers to come on board easily and economically, so they need more inclusion now, at the start.
Simon Lloyd-Evans
CIO
Bupa Care Services NZ
 

Hi Graeme,
This is an exciting plan and thanks for putting it up on The HIVE and collecting feedback this way.  Allow me to use a small point as a wedge to a larger one... The plan talks of sharing 'Patient Vitals.'  In the cardiothoracic surgery example the words are flipped and it's 'Vital Patient Data.'  Either way, the word choice conjures the idea of 'vital signs' (heart rate, etc.) but in fact the plan indicates that it includes current medications, current problem list and possibly lab tests and historical data.  So, it seems what we're actually talking about is shared access to one or more repositories collected from community (and maybe hospital) health transactions.  This sound far more useful than sharing vital signs, but also more imposing for at least some citizens.
What if the example, instead of cardiothoracic surgery, were type I diabetes or epilepsy.  And what if the patient in question was a 17 or 18 year old female who has prescription birthcontrol or has had an abortion, and perhaps (I realise I'm piling it on) has divorced parents who have differential knowledge of her sexual activity.  Many GPs keep their screen visible to the patient and anyone accompanying the patient to a consult; which is great - but in this scenario, a routine diabetes check might show a surprising medication list.
I reckon it's all manageable.  And I'm a huge proponent that we don't throw the baby out with the bathwater.  In my experience, the vast majority of older adults and people with significant health conditions are highly in favour of effective information sharing and seldom have any qualms with complete disclosure to providers.  The potential benefits are huge and it's our duty not to deny the benefits to that large group due to edge cases.  However, for just that reason, we need to actively probe the edge cases and find out exactly how to operate the system in a manner acceptable to the population that it will cover.
Moreover, note that even the Phase 1 community e-pharmacy will already invoke signiicant data sharing issues... we don't need to wait for shared care.
I had noticed in the Shared Care Planning workshop some weeks back that the consumer and clinician views on information sharing were unaligned.  It's fine (necessary) to have a consumer forum and to have a clinical leadership group, but we need to huddle doctor, nurse, pharmacist and consumer together and work through mutually satisfactory modes of operation for the range of consumer situations.  We also need to be sure that pilot studies are well 'instrumented' with qualitative feedback to learn how people make out in the real world with this information sharing.  I believe it goes against human nature to evaluate these situations well as hypotheticals (although we should give it a try before piloting).
I'm also receptive to Simon's point above.  The elderly are such major health consumers - let's not squeeze them and their carers off the stage.
Exciting times ahead.
 
Jim Warren, Professor of Health Informatics, NIHI Chief Scientist
The University of Auckland 

Nathan, thank you for this feedback and acedemic references - very helpful...I have attempted to provide an initial answer for each point. Please do not hesitate to respond if I have missed something.

Obtaining data for the National clinical data repository

If I have understood your point here, it is that a hospital cannot be seen as a single entity with information only capture on patient entry and exit. As you point out, within a hospital 'campus' there are a range of services available. There needs to be visability when a patient is transfered between hospital based services. In response, we expect the clinical data repository to capture key health events, however you have reminded us to make sure the 'events' within a secondary care environment are also recorded in order to capture a full history of activity.

Managing the data within the National Clinical Data Repository

The way we have been thinking about your concept of an 'Intelligent Agent' is that key items of health information need to have a standard format, 'look and feel' and operation (no matter what information system or web site a user is accessing the information through). Examples would include access to medication information, labs and radiology information, or patient vital information (I note we may need to change the use of the word "vital" after the recent HISO meeting - see the comment from Jim Warren!)

Implementation issues

Your comments in this section are music to my ears and describe much more fully the basis of the approach being taken by the National Health IT Board (and in fact the National Health Board in transforming the health system). The words in the draft plan do not do justice to the leadership from Dr Murray Horn and Chai Chuah. Much of their thinking is influenced from the case studies presented by the Institute of Healthcare Improvement (IHI). It is all about clinicians committing to improve quality outcomes over time, by defining a population, applying a scientific approach to testing practices,  measurement, transparency and knowledge development (and published). With this context the Health IT Plan must deliver on a dual set of objectives: support and enable clinician led groups following the above approach, while also creating a health information platform for the wider sector (and in this way guard against diffusion).  

   Mobile Health

Mobile health has been in our thinking from day one. I agree the current draft is silent on the subject. It will need to be addressed in the final plan. The overall design we are working to is to deliver a core summary 'Shared Care' record that can be used as an authoritative information base by independent health applications whether web or mobile based. This approach will promote ongoing innovation in the Personal Health Record space.
Thanks again for your comments...

David, thank you for taking the time to review the plan and provide feedback. I will ask the board to consider very carefully your suggestion about the title "Enabling an integrated, personalised healthcare model". I particularly like the point you made near the end of your comments "We should however be held more accountable for leading this change than we currently are. Likewise we should take more responsibility for assisting your Board in implementing the HIT plan." The Board is leading a whole-of-system approach to improving the use of health information in New Zealand, however we are very clear that lasting solutions will come from clinican leadership in local, regional and national settings. We are happy to receive your feedback from Boston and we look forward to engaging with you on your return.

Hi Chris, thank you for a very good summary of some of the challenges in the public health domain caused by the current 'siloed' information systems utilised across the health sector. I took from your comments that you felt public health was out of scope. This is not true, the scope of the plan is whole-of-health-system, however we have strong view that capturing personalised health information during the patient journey creates the foundation for all other uses of information including: policy, planning and funding, research and ... public health. Some areas of improvement for Public Health I would expect from the plan are: common data definitions, better access to information and additional alerts triggered by key events. I would very much like to hear your thoughts how public health responsibilities could be supported by the national view of clinical information captured in the proposed regional repositories.

Thanks Simon, we have received some comments along this line...better to receive them now while we still have a draft plan! Please be assured that we have thought about health care delivery and information challenges relating to aged-care. The role and support of NGOs is also being discussed at National Health Board level.
The basic premise of the approach we are presenting is that core health event information needs to be collated and presented as authoritative (and trusted) information to patients, clinicians, the health workforce and the patient's carers (including NGOs and family etc). Noting at this point - the way this information is presented needs to have appropriate safeguards. Now, in the next 2 years, the plan sets out (amoung other priorities) the importance of implementing a standard list of medications. We strongly believe this will have a very positive outcome for patients receiving aged-care services.
The other point you make is the challenge of affordability of information solutions in terms of NGOs. To be frank this is an issue for the whole health sector not just aged-care and NGOs (although it is more apparent). As a small and comparitively modest 1st world country, in terms of national income, New Zealand must ensure that every investment in health information solutions fits into the whole. NGOs need to work together, share resources and build on the success of other NGOs in creating low cost and effective information solutions. An example of the latter is MMPO in Christchurch.
 
 

Jim, I read your comments with interest. I have been waiting for someone to test the area of access to personal health information. I think you have been very thoughtful in the way you have introduced some of the real life challenges that arise when sharing personalised healthcare information. Rather than comment directly on your ideas (partly because it is getting late!) I have decided to state some prinicples and actions we are taking to engage and find a suitable landing point in this area.
1.Privacy - the legislation that supports privacy in NZ (while currently under review) is very effective in protecting the rights of the individual. The key 'rule' is that when information is collected the organisation collecting the information needs to divolge to the individual what the information will be used for. This is a sound prinicple. We will work with healthcare organisations to make sure there are common practices in the this area.
2.A prinicple of clinical professionalism - put simply the plan relies on clinicians and health workers to use their professional judgement at all times when they are sharing health information.
3.Consumers must have confidence that sensitive health information will be protected (and managed appropriately.
4.The board is running 6 regional consumer forums in July/August. The forums will have braod representation from consumers, however we will make sure the young and the aged will be well represented. The idea is to have an initial discussion in a workshop format with an opportunity to engage with clinicians in the latter part of the event.
Thanks for your comments.

I certainly agree Simon that this is a real area of concern.  These are high needs citizens, with comorbidity and at high risk of showing up at emergency departments in the middle of the night.  From what I've seen at a number of facilities it is still largeley paper based records.
I'm the sole visiting GP to a psychogeriatric HDU.  I have enrolled one of my residents on medtechs managemyhealth, and made this available to her daughter who has EPOA in Australia.  It has been very useful keeping her daughter up to date with diagnosis, medications and lab results.   This only works because I use remopte access to my surgery from the facility and run them all on medtech via my laptop.  Clearly this doesn't work in a paper only environment.  (I copy off the consult notes for the paper precrd in the hospital).
There are a number of the business cases for Better Sooner More Convenient which are looking to use managemyhealth as an interim step.    It may well be we need another discussion on hive to talk about this small part of the overall vision for an EHR summary.  Making the record available to ED and other appropraite health workers is the aim.

Hi, many important issues so far have been identified and discussed - very valuable contributions. However I'd like to point out to two issues which I believe have not been touched:
1) Need to agree on a simplified set of clinical workflows around major medical conditions - mostly chronic diseases. While the shared 'care' will be based on evidence based clinical pathways I think the administrative and clinical workflow should be aligned with a core set of national/regional workflows crossing the whole continuum of care. Naturally IT will shape around existing workflows but at times we may want to reengineer certain processes to fit IT better.
This is an emerging issue as we get started started with ASSET Project - even if the handling of the medical conditions are equivalent, the people and processes differ - albeit slightly. But this may hinder some of the abilities we would want to enjoy using IT. So I advise to think about 'core clinical workflows' together with 'core data sets'. Map of Medicine on the clinical side and IHE profiles on teh technical might be good resources.
Agreeing on such common processes may also make possible simple but essential decision support such as warnings, reminders or treatment adherence etc. I suspect the 'care plan' part of the shared record will also rely on this.
2) Must think on the requirements of clinical research (other than public/population health); especially using live clinical data for such purposes. This is perhaps future-work but it'd be wise to consider the type and granularity of data to be held in shared meta-data repository and clinical data repositories accordingly.
Best regards and good luck!
 

Thank you for this opportunity to reflect on the National Health IT Plan and discuss aspects with you Graeme. Well done on an ambitious and yet sensible plan for implementing an already ambitious strategy.
 
I would like to bring to your attention what I believe make up the Science and Art of Clinical Care in a person centred health system. In terms of the Science of Clinical Care, we’re doing well: evidence based care, research outcomes, and the collection of copious amounts of data in our health IT systems all over New Zealand. The quest for a central summary EHR is an important part of this Science, forming a core set of data as a starting point for every episode of care a patient may need.
 I’m not convinced that we’re doing all that well in terms of supporting the Art of Clinical Care. In IT terms, I would suggest that the key features of this Art include the following. 
(1)   Mobility of both healthcare personnel and patients. For example, a sixty-year-old person with diabetes and complications probably sees her GP the most, but also visits the specialist, gets meds, has her feet checked and sorted by the podiatrist and periodically needs a district nurse to sort out chronic wounds on her legs. Each person she sees is disconnected from the other. Even in a hospital where the most sophisticated IT has already been implemented, IT is tethered to desks and usually more difficult to use than pen and paper – we’re competing with the mobility of a small notebook and pen in a nurse’s pocket. I believe that the National Health IT Plan doesn’t adequately address this issue and should at least indicate the baseline for mobility and the untethering of the technology.
(2)   Healthcare intelligence/knowledge management. I won’t bore you with definitions other than to say that you know you’re looking at healthcare intelligence when you see a clinician making a good decision upon which to act and the result makes a difference. This is an art and data/information doesn’t do it for me. Despite all the good things our health IT does for us, we’re still entering data into a databank (no, not a database anymore) and trying to use data for the Art of Clinical Care. The Art is more about the people than it is about the technology or information. It’s more than good change management. It’s about HOW people use health IT and the results of that use. It’s a big step above repetitive data/information use. It’s not clear in the Plan how intelligence is supported.
(3)   Shared care aka interprofessional practice aka multi-disciplinary care aka team care. We’ve been struggling to achieve this for centuries! We’re good at sharing care in a linear fashion. For example, a GP refers a patient with a skin lesion to a dermatologist who discovers cancer and refers the patient to an oncologist who hospitalizes the person and operates and returns the patient to her GP after a clinic appointment for follow up after discharge from hospital. In other words, a clinical scope of practice defines which type of clinician is accountable for which aspects of care a patient needs. The Plan’s signal that shared care is Phase 2 is laudable but without policy and changes in the education of clinicians we’re at risk of making the tail wag the dog. I realize that the Plan is being developed in close communication with the Healthcare Workforce Team and a group of clinical leaders, but it would be useful to overtly link the two together (and other influencing groups/stakeholders) so that the people who are expected to make the changes required for shared care know what’s on the To Do List.  
There’s a lot more to the Art of Clinical Care… I’ll leave you with the suggestion that a To Do List that’s drawn up in collaboration with clinical leaders (not limited to doctors and nurses) and policy makers and educators will help us to make the changes we need to support the success of the National Health IT Plan. 

Graeme,
 
Thank you for taking time to respond to my initial posting. There is nothing more to add to the comments made already.
 
However, there was one outstanding issue that I also felt has not been adequately addressed or discussed within the proposed plan. I understand the deadline for comments has passed but I wanted to see what others thought of this issue.
 

Health Practitioner Access to patient information and mobile privacy issues

 
Background:
New Zealand is far more advanced than other countries in establishing a common currency for health IT users online.  With both the National Health Index number to identify patients and the Health Practitioner Index for details of health professionals.  However it does not appear to be utilising this data to control access to information or functionality within Hospital Information Systems. 
 
While working in the U.K. at Chelsea and Westminster hospital (using Lastword an IDX solution that was taken over by GE Health and adapted into newer version Centricity) we were able to set up different levels of user functionality based on clinical level.  E.g a booking clerk could not check patient's biochemical results and a clinician could only prescribe drugs for a patient under their care.  This helped improve security with the actions of users being controlled and monitored.

Question
What are the possibilities of establishing an national directory or practitioners based on HPI numbers to serve as an "active directory"( http://www.microsoft.com/windowsserver2008/en/us/ad-main.aspx ) ? this could be based on a share point concept and used to control access to information.  This will be particularly  important with both patients and community based health care workers wanting to access this information out of the secure hospital environment, maybe on a mobile phone or iPad.
 
I guess what I am trying to get at is; are there going to be any plans to establish role based rights management rules for patient data viewing, editing and writing?
This is fundamental if you are going to look to establishing nationwide interoperability  for things like electronic prescribing and control the access to sensitive patient information within the national clinical data repository later on.
This concept of role based access and a ingenious solution for mobile access to classified patient information was described by Krawczyk (2005) in his masters dissertation.  He also highlighted a potential way to make this sensitive health care information secure within a cloud like environment. This can be seen in figure 1 below.
 
Figure 1: Proposed system for tablet PC multimodal validation (Krawczyk 2005, p.14)
 

It is a tricky area, confidentiality. I appreciate it falls outside the scope of the national health IT plan and already regulated by the privacy commisison.
 
However the need for role based control and access to information will only become a bigger thornier issue unless it is nipped in the bud earlier on. 
 
Thank you once again and it would be interesting to see if others see any potential to this idea.
 
Many thanks,
 
Nathan Billing,
Health Informatics student

References 

 
Krawczyk, S., 2005.  User authentication using on-line signature and speech.  M.Sc.Michigan: Micigan State University [online].  Available at http://biometrics.cse.msu.edu/Publications/Thesis/SteveKrawczyk_VoiceSignFusion_MS05.pdf  [Accessed 21 May 2010].
 

Feedback on the draft Health IT Plan is now closed. Thank you to everyone who contributed their views. Please continue to use this thread to raise any further thoughts or questions you may have.
Next week we plan to run new threads on common themes identified in the feedback.

I was pleasantly surprised to see such a focus on client access to their record in the HITBplan, and was particularly interested the read about the uptake of the HealthSpace programme in the UK. The following is a quote from the site below which has some details of the report by University College London on the summary care record scheme   http://www.computerweekly.com/blogs/tony_collins/2010/06/highlights-of-confidential-ucl.html 
the HealthSpace programme, which was built on the assumption that a significant proportion of patients will have the motivation and capacity to 'self manage' their long term condition using this technology; that this will reduce costs to the NHS; and that patients' access to their SCR via HealthSpace will contribute substantially to improving data quality. The findings of this study - that few people are currently interested in using HealthSpace to manage their illness or access their SCR - suggest that it is time to revisit all these assumptions."
The report makes interesting reading with some important lessons for NZ.  Another classic is this:
 "The SCR and HealthSpace technologies contained a number of inscribed assumptions about the nature of illness and the behaviour of patients and staff. For example, inscribed in the SCR was the assumption that GP practice staff would enter all key data in coded fields on the local record...Mismatches between these inscribed assumptions and the reality of clinical work or living with illness explained much of the non-adoption, partial adoption and abandonment of these technologies at the level of the individual user."
do have a read - very worthwhile -and lets try not to make the same mistakes.
Maggie Broome 

There is some really good reading at this website regarding a report on the UK efforts with the Care Summary Record and the patient portal.  Lots for us to learn from, particularly around assumptions about both clinician and patient behaviour.
http://www.computerweekly.com/blogs/tony_collins/2010/06/highlights-of-confidential-ucl.html
well worth a look.
Maggie