Health Transformation

CE-classification in Medical Information Systems?

What is a medical device? In MEDDEV The EU Commission try to sort this thing out. It is not easy, according to the rules, to know the borderlines.
In Sweden the Medical Product Agency just released a document proposing how to classify software in healthcare.
This report has the conclusion that software, direct or indirect, built to influence diagnose, care and treatment of a single patient must be regulated under a product safety framework. The authors have not found any other suitable framework than the one for Medical Device.
The tree directives in EU are:
o Active Implantable Medical Devices Directive (AIMDD)
Directive 90/385/EEC - OJ L189/ 20.7.90,
o Medical Devices Directive (MDD)
Directive 93/42/EEC - OJ 169/ 12.7.93
o In Vitro Diagnostic Medical Devices Directive (IVDD)
Directive 98/79/EC - OJ331/ 7.12.98
Standards helping along the path include ISO 14971, IEC/TR 80002, IEC 60601, IEC 62304 and IEC 62366.
Software used in healthcare is a “product” according to the authors. They think that it must be looked upon as a product and terms like “project”, “service” etc must not be used describing a medical information system. If you see it this way, the Medical Device Framework is ready to use. The authors see in the future that medical software, not CE marked, should not be chosen in public purchase.
Is this a feasible path?

Wash your hands of IT

Once again the public opinion has been grabbed by life-threatening circumstances in hospitals. This time it was television in the Netherlands that did some undercover cleaning work and found out that patient televisions had not been cleaned for a long time and that blood on the operating theatre lights could not be wiped away properly, as another patient was already waiting. Infection prevention specialists are stunned, because procedures are widely distributed across hospitals and specialized cleaning companies.

Washington Post - Incorrectly Infers Impropriety by HIMSS & HIT Lobbying

Yesterday, Saturday 16 May 2009, the Washington Post went looking to create controversy where there really is none. Their position is that too many HIT vendors stand to profit from the Obama Stimulus Package (i.e., ARRA HITECH Act) and that HIMSS lobbying was at the root of this evil. In fact, they are on the wrong side of the argument. In fact, I believe that they should have been challenging the fact that no one was listening to the lobbying before. What's your view?

The article (http://www.washingtonpost.com/wp-dyn/content/article/2009/05/15/AR2009051503667.html) starts with the following text:

"When President Obama won approval for his $787 billion stimulus package in February, large sections of the 407-page bill focused on a push for new technology that would not stimulate the economy for years.
The inclusion of as much as $36.5 billion in spending to create a nationwide network of electronic health records fulfilled one of Obama's key campaign promises -- to launch the reform of America's costly health-care system.
But it was more than a political victory for the new administration. It also represented a triumph for an influential trade group whose members now stand to gain billions in taxpayer dollars.
A Washington Post review found that the trade group, the Healthcare Information and Management Systems Society [HIMSS], had worked closely with technology vendors, researchers and other allies in a sophisticated, decade-long campaign to shape public opinion and win over Washington's political machinery."

Oh my! A group looking to improve healthcare through increased use of information technology...shear scandal. Clearly HIMSS and the Obama Administration must be wrong to promote moving health into the 21st century through the use of technology and bring healthcare in par with the service expectations of the general population and the use of technology in other industries is simply beyond reason...right?

What are they thinking about?

The article is absent of any relevant research about the issues of HIT, health interoperability, and the relevant benefits. There is no mention of the fact that most every government in the world is investing in the integration of health information. There is no mention of the fact that there are numerous cases where HIT is in fact improving the quality of care and while reducing the cost of care.

Anyone who has been in the industry knows that the use of HIT has been pursued for many years. We know also that the US is one of the few countries where the majority of the primary care providers (>80%) are providing care WITHOUT systems. These physicians have no systems to handle such basics as electronically recording the care provided, alerting doctors of conflicting treatments or prescriptions, reducing prescriptions errors due to poor hand writing, or simply reducing administrative cost associated with copying/sharing medical records. Of course the Post did not bother to consider these current condition.

HIT is surrounded by a vast number of complex issues, but the Post is simply proving again that they are just out to sell papers by raising a non-issue for the sake of creating controversy...not inform the public and support a change that must happen.

Personalized solutions in a standardized clinical environment – a paradox?

In a future scenario with implemented standard processes and clinical pathways is there any space left for the individual? Quality index, cost control, effectiveness and a common information model and structure does it stop the individual initiatives and creativity?

Today most information is accessed from a variety of “siloed” and disparate systems. The evolution points out standards have a greater impact now and a lot of good work is being done on interoperability.

In an environment with several hospitals and primary care units etc. with the same owner, public or private, the issue of common enterprise-wide data is becoming more important. Other sectors like Finance, Telecom and Retail have already experience of this, Master Data Management.

Community EHR/HIE Adoption

Over the past few months, as ARRA HITECH Act has begun to stretch the imagination of more people, the desire to implement increasing numbers of interoperable systems or information exchanges has increased dramatically. That's great, but let's not forget that adoption by the physician/clinician remains paramount. Oh, and by the way, the challenge in the US is complicated by extending the implementation out to the primary care doctors 80% of which have never had a EMR. So some quick thoughts on paying attention to the issue of adoption:

The solution needs to be 'meaningful' to day-to-day activites (not just in terms of of what the government wants to see before they pay out the incentives).

An empty EHR/EMR is not really going to be helpful so rapid loading of current patient data will be essential to delivering value.

If there is some automation in place, you probably ought figure out how to use that in the overall solution because rip and replace is a flawed strategy (ask some of the participants in England's national program).

Dropping a solution into a practice is only the beginning. Support service needs to be included.

Clinical decision support will improve effectiveness of the practice and will become an essential requirement for improving experience for provider and patient.

Workflow support will improve efficiency for the practice.

Together, the implementation of these capabilities will help to streamline the challenges of adoption.

NOTE: At HIMSS, there was a common fear that the new stimulus money will bring in new IT service providers and consultants who don't really understand these issues. If you are in the market for consulting assistance on these fronts, buyer beware.

HIMSS, Chicago 2009 “Transforming the Industry Together”

Visiting the annual conference and exhibition some reflections pops up.

• Clinical decision support
• Healthcare processes
• Workflow
• Unnecessary errors/mistakes in healthcare
• Patient Health Record

In World of HealthCare Congress – 2005 in Les Fontaines, Paris, Per Båtelson CEO Capio Group Sweden, talked about “Doctors should be like pilots, not artists” in the context of decreasing errors/mistakes in healthcare.
We have heard about horrible errors/mistakes in healthcare at WHCC and HIMSS every year there after. This year at HIMSS we heard Dennis Quaid, CEO Quaid Foundation, talking about his, and his family, experiences in healthcare.
The work, being done and the ongoing, about clinical pathways, process models, workflow and clinical decision support clearly points out the need and stress the fact that the number of unnecessary errors/mistakes must decrease.
As a complement to this the Patient Health Record likes HealthVault, Revolution Health, Google Health etc. engage people to be active in the process and able to follow the activities made by the Caregivers. Compare to the banking industry and the evolution of the internet bank.
When do we see the curve of unnecessary errors/mistakes in healthcare decline?

Changing the tune of the discussion

Moving away from functional and non-functional requirements of a system is probably the top priority in health informatics at this moment. "Ask not what the system can do for you, but what you can do with the system" seems an appropriate statement in these times of change.

Over the last few weeks I have been engaging heavily in discussions with clinicians, health IT managers and managers of healthcare provider organizations. The topic was invariably on the Electronic Health Record (EHR), beit a regional health information exchange or a local EHR system for one institution. A common theme in these discussions is the necessity to take the perspective of benefits to the individual healthcare professional, to the patient and to the health system at large. Participants often feel they need to express their requirements of an EHR system in terms of functionality, security and performance, rather than focussing on the ways in which their daily lives can be improved using such a system.

Attribute Based Access Control

Access control is one of the key issues handling the information flow inside healthcare actors, between healthcare actors and cross border healthcare actions.
Conventional access control models are for example Identity Based Access Control (IBAC) and Role Based Access Control (RBAC). In IBAC access permissions are directly associated with a subject and it is difficult to scale. In RBAC access permissions are based on the role(s) a subject is performing. RBAC gives better scalability and ease of use but have drawbacks.
In Sweden a national security application are under way and is built upon Attribute Based Access Control (ABAC). This model consists of three groups of attributes:

• Subject Attributes
Associated with a subject (E.g. identifier, name, job title, role….)
• Resource Attributes
Associated with a resource (E.g. metadata elements …)
• Environment Attributes
Describes the environment or context (E.g. current date, time, classifications…)

This Swedish initiative will be the world’s largest XACML 3.0 (OASIS) deployment.
I see the beauty of policy’s and the abilities in the Service Oriented perspectives but in the daily caregiver situation could the complexity making the policies be too high and leaving doors open accessing confidential electronic patient records?

What Does American Recovery & Reinvestment Act mean for HIT?

Clearly US Stimulus Package for HIT ($19.2B) shows there is a strong desire to get codified patient data moving in an electronic form, but why is it going to work? Are providers really going to be interested in letting that cost savings impact their share holders? Will the short-term incentives really going to allow providers to see a new value proposition emerge where they can start new revenue streams (e.g., telehealth/telemonitoring)? How does anyone limit their spending of Stimulus money on the mandated migration to ICD-10 and ANSI X12 5010 transactions? Frankly, it’s too early to tell, but it will be interesting to start hearing what folks are planning. HIMSS could be especially interesting this year…maybe not the presentations or even the vendor booth…

One other thought on this topic of why is the Stimulus Package for HIT a good idea: has anyone thought that maybe there are other purposes for this increased flow of computable health data? While there are obvious benefits for the pharmaceuticals, the opportunity for more complete evidence based analysis over a larger population will likely give all-seeing payers the ability to identify which treatments are yielding the best clinical results. Will that lead to a new way to influence care? Will Medicare and Medicaid continue to pay for treatments that are not as effective as those that are deemed to be the best? Is the Obama Administration that clever?

Keep watch for those unintended consequences, that’s where it will be interesting. In the mean time, don’t hesitate to propose for those grant moneys and loans when they come available in 2010…assuming the economy has turned by then.

Patient privacy between politics and practice

In all Health Information Exchange initiatives I've come across, discussions about patient privacy are taken very seriously and rightly so. Who needs to have access, how to manage patient consent, when to restrict patients' access to their own records, all these issues are taking a long time to resolve. However, current live experience with patient privacy in practice sheds a quite different light on these discussions and makes you wonder whether the often political discussions are focussing on the right issues.

Current practice in treating cancer patients involves local, regional and national clinical networks and both local and specialist multidisciplinary teams (MDT). When a suspicion of a fairly rare type of cancer surfaces, clinical guidelines require the GP or surgeon in question to contact both the local and the specialist MDT to discuss the diagnostic results and treatment options. This means, among others, that diagnostic images are shared with a specialized cancer center and the patient may be referred to the specialist surgeon for further diagnostics and treatment.

Upon arrival in the specialist center, I was quite surprised to see the MRI images from the regional hospital already loaded in the local PACS system and available to the specialist surgeon, especially when I saw that the spelling of the patient's last name was all screwed up. Of course, in this case, the error was easily corrected, but it made me wonder.

How come we are discussing all sorts of complicated privacy issues about sharing patient information when in practice doctors and patients know quite well when and why pertinent data needs to be shared? How come, at the same time, simple issues like common and safe patient identification have not been realized yet? Should we focus more on the basics and tackle the more complex patient privacy matters in a more restricted clinical context, like cancer care? Both political an practical experiences are more than welcome!

About the authors