The critical need for open-source health care

One of President Obama's biggest presidential ambitions is to reform the U.S. health care system. With more than $2 trillion spent each year on health care costs, an estimated 25 to 30 percent of which is administrative waste, one of the best stimuli to the U.S. economy could be to fix our broken health care system.

It's unclear, however, whether the Obama administration plans to tackle one of the root causes of U.S. health care inefficiency: closed, siloed, and payer-centric data.

Talking with a neighbor the other day who specializes in health care IT, he related some unsettling facts:

Picturing patient data interaction

• In today's system, information/data ownership drives profits for payers (insurers), software vendors, and major hospital chains, but doesn't improve efficacy or efficiency, and both doctors and patients are largely left out of the information loop with little or no access to patient data.
• Fifty billion health care transactions are processed each year to describe who will pay whom, but very few of these actually describe the efficacy of treatment.
• With over 1.1 billion doctor visits, 2.6 billion prescriptions, and millions of lab tests and imaging (x-rays, CAT scans, MRIs, etc.) every year, the amount of clinical data generated is too large to be handled on paper--and currently only 34 percent of these billions of pieces of information are transmitted totally by electronic means.
• Although doctors and hospitals have been installing computerized record systems at a growing rate, almost all of them deal with billing, not clinical or diagnostic information--and even if they do store diagnostic information or prescription data, different providers use different--usually proprietary--systems that cannot talk to each other.

In other words, poor capture and reuse of health information is one of the main reasons U.S. health care costs more, for poorer results, than almost any other industrialized country.

This is where open source could help.

OStatic talks about a few possible solutions, but I think our central need is for open-source middleware that connects the disparate systems and standards that make up the U.S.' Byzantine health care system, and focuses data on patient care, not payer care.

Open source is perfectly suited to this sort of disjointed system. An open-source project has the potential to provide the security and standards currently lacking. And while the U.S. federal government is starting to make some strides with systems like Connect, I suspect we'll find that a project outside the government will tend to work more efficiently than one designed and driven by the government.

Government, after all, is not really the solution to the health care mess. Coordination of disparate data sources is, and that's the sort of complexity that market-driven open source thrives at solving.

Just look at Eclipse, Firefox, Linux, and other leading open-source projects: like free-market capitalism, they effectively manage the coordination of widespread, competing resources to produce projects that benefit a host of different participants.

I think there's an opportunity to accomplish this same thing in health care. We have some examples of successful open-source health care systems, like Medsphere's adoption of the U.S. Veterans Administration's VistA ERP system, but we need more.

Open source arguably can do more to help improve the U.S. health care system than any policy President Obama can enact, no matter how well-intentioned.

Follow me on Twitter @mjasay.

[Get_Bent]

The problem isn't lack of open-source software; it's the lack of open standards for data interchange. If there were some clearly defined formats for exchanging medical data among disparate systems, it wouldn't matter who made the software.

[SIGHUP]

The problem is not a lack of open standards for data interchange, its that there are too many of them. Everyone does something different. There is a saying that goes something like the best part about standards is that there are so many of them.

[Orion Blastar]

http://www.openemr.net/what_is.jsp for starters. It is OpenEMR software for exchanging medical records, most open source healthcare software uses that standard.

Check Wikipedia for a list of more open source healthcare software:
http://en.wikipedia.org/wiki/List_of_open_source_healthcare_software

[rapier1]

There is an open standard. Its called HL7 and, at least in the US, is widely adopted. The problem is that HL7 is a freaking disaster. A classic example of a protocol written by committee. There are multiple incompatible versions and every implementation seems to have its own idiotic quirks.

As a side note; there are some good open source implementations of HL7 translators and message routers available. Mirth is a great example. Of course, its not the best implementation.

[42istheanswer]

The problem is basic human greed. No one is going to come up with a system, open source or proprietary, that will deal with that. This is all an exercise in futility. Spin that wheel faster, my friends.

[LeeMD]

This is right on the money. "closed, siloed, and payer-centric data" is the problem today. Obviously privacy concerns will also hamper the widespread adoption of open standards.

[jtwiles]

Working in healthcare decision support, I can attest to the focus on ?payer care? over patient care with regards to data. However this is not a testament to lack of open-source middleware but to the fundamental flaw of a multi-payer system; that is that payers do everything they can to pay as little as possible and hospitals (needing to at least break even in order to stay afloat) try to get everything they can from the payers. No wonder most of the software focuses on the financial side of healthcare. Open-source middleware will do nothing to solve that fundamental problem of each side trying to get one over on the other in order to make money. I agree completely that coordination of disparate data systems is greatly needed to improve healthcare reporting and patient results, however just saying ?open-source? yeaaa! that will fix it, isn?t going to do jack. Medsphere is just another software company competing against giants like Mckesson. It?s hard enough to find skilled IT professionals who have the health care knowledge necessary to be effective at something other than just installing software and building networks. Start throwing in a bunch of open-source software in the middle and you introduce a whole new layer of complexity in a field that is already lacking in skilled workers and overflowing with complexity. We desperately need universities to start offering health information science majors/minors intertwined within computer science programs. This way we have kids coming out of school with healthcare information on the brain. Right now we have lots of IT people and we have lots of Healthcare people and very few people with both. Just my 2 cents.

[Orion Blastar]

I worked for a medical software company who used close sourced standards for medical tray inventory and sterilization software. They had competitors and had to hack their competitor's databases to access the data and import it into their own software. While many companies write software to do the same sorts of things, their databases are different. Most of their competitors used flat files or DBF files, and they used MS-SQL Server 2000 and MS-Access. I had helped convert their software to MS-SQL Server 2000 and MS-Access from MS-Excel files.

It has been since 2002 since I worked there so my five year no-compete contract is up. I could write open source software to do the same sorts of things they are doing, but with a different database. I wanted to do a Linux solution with MySQL or PostgresSQL and use PHP to serve web applications so that any device with a web browser could access the data.

I lack the money to develop such a system, and I would need to do research with hospitals to learn how their inventory and sterilization functions work, and build a new database around that. I don't want to copy what I did for my former employer, I want something totally new and written from scratch. One that they cannot claim I copied from them.

[rapier1]

There are open source middleware solutions for data transfer and integration. That's not really a problem. HL7 is a mess but, for better or worse, its a mess that enough people understand to make it a functional solution. The problems are
1) Physicians are not convinced that this is a solution for falling revenue. It is - especially when used in conjunction with a medical decision support system - but they aren't convinced of that and without physician buy in you do not have a solution.
2) HIPPA creates a policy level firewall that hinders the free flow of data. While I understand the patient's desire for privacy regarding their medical data it is a hinderance to the actual task of moving data across organizational boundaries.
3) Practices generally make conservative investment decisions. No one wants to be the first to put their toe in the water and even open source won't help with that. Few OSS solutions are going to be plug and play and therefore require what is seen as a significant IT investment. They also view OSS as riskier than closed source solutions. They'd much rather pay an annual maintenance fee to GE or Cerner than risk having their investment abandoned and their data stranded.

[PACSferret]

There are a handful of reasons why IT has had poor impact in healthcare (BTW certainly not just in the US):
1. Health is, and should remain for a very long time, very human-centric. Whether one talks about doctors, radiologists, nurses, physios, dietitians or any of the other specialists (let alone patients and relatives), it is very different to industries where 'transactions' are much more discrete.
2. The major incumbent vendors have no interest in sharing information - the major vendors in particular who sit on the commitees that produce the standards like HL7 and IHE and who sell components that do 'everything'
3. In reality, information flow between specialties is limited. The physio doesn't need to know everything the intensivist knows - and since they both have to be in the same place as the patient - why complicate the process with technology?
4. Decisions are often made in a very parochial way - very focussed on specific specialties and departments.

That said, is being made gradually, although not being led by the US.

[Ian Skerrett]

Matt,

You should take a look at Open Health Tools. They are modeled after the Eclipse Foundation and have some of the big healthcare providers (VA, UK NHS, Canadian Infoway, etc.) involved. As far as I can tell they are the closest to achieving the collaboration you are lookig for

http://openhealthtools.org/index.htm

Ian Skerrett
Eclipse Foundation

[PACSferret]

A case in point:
http://www.nytimes.com/2009/03/06/opinion/06coben.html?scp=7&sq=medicine&st=nyt
Hey I'm an IT pro & software dev for almost 30 years & I agree with her. I've long been of the opinion that some way 'outside-the-envelope' innovation is needed & perhaps (just perhaps) open source has the wherewithal to deliver.