Open Source EMR vs. Commercial EMR

The 'open source' EMR debate is one that has raged on for years. The concept is quite straightforward, open the code for a particular EMR system. Allow end users to modify and adapt the program so that enhancements can be progressively built into the EMR system. Share the experience of the end-users with others in the open source environment. Progressively the software will improve and will achieve greater functionality. More important, through the participation of 1000's (hopefully) of users, there is reduced likelihood that the software would become outdated and that the vendor may fail. At least that is the theory.

One of the greatest success stories in the open source movement has been the Linux operating system. Through the input of hundreds of thousands of users and technical wizards, Linux has developed into a stable and inexpensive alternative to MS Windows. However, I am not sure this process is applicable in the health care sector, especially as it applies to EMR. Firstly, the scale is different. In Canada there are less than 70,000 actively practicing physicians. Even in the US, where the numbers are approximately 10 times greater, the number is small in comparison to the effort that was applied to Linux. Of the 70,000 physicians in Canada, most are devoted to their clinical practice. Few have the time or the skills to become effective contributors to the development of software products.

So my question is this! If open source EMR is the answer, how do we achieve a critical mass of end user/developers who have the skill set and time to ensure the ongoing development of the EMR, particularly as the needs and demands of the end user increase with the growing numbers of interfaces to labs, diagnostic imaging and hospital systems? I have visited some open EMR user forums and even as a tech savvy physician, I have difficulty understanding the technical jargon relating to implementation and system debugging.

If any physician has thoughts on this subject or has implemented an open source EMR, please share your thoughts by clicking on the 'Comments' button below.

Comments

[peter richards]

but whether its open source or closed source, it's likely that you will get all the technical and training support you need. If you can do it yourself in open source then you can do so, as I have with the latest OSCAR which I now use in my office at zero cost but I could pay $3000 per year for install, training and support if i needed it. And if the company goes out of biz as they have in the US and UK or the company is sold to a bigger company, you have a big problem. With open source, it's likely given the tools that you get your data out.

[Morgan Price]

Alan,

I think you have hit the nail on the head as to the perception of why an open source project will fail in Canadian healthcare. Remember that Open Source begins as a license, nothing more. Instead of selling the software, you can sell support and further development of the product.

The lack of developers argument can be used for closed sourced EMRs in Canada too and yet there are about 75 EMR vendors in Canada... There are many many challenges with Open Source that parallel a closed source project (dollars, skills, time, coordination, design, etc). I do not think that it is necessary to have such a mass of developers to create a good tool, open source or not.

Where, I think, Open Source has some benefits are:

a) User Has Control - as Peter mentioned, you CAN get the product for free, or you can chose to get support (a wise move unless you're technically oriented like Peter). Having access to the code does mean that you should be able to add or at least access your data. Should there be an issue with your support provider, you can change providers. This is particularly important in healthcare where, currently, it is not an easy thing to move your patient data between any systems.

b) Sharable Resources - Outside of the open source EMR community (Which is very small) there is a large collection of open source tools which are free to share -- developers can leverage this wealth of features and use them in an EMR, some iwth very littel modification. With a good architecture, this is a way of rapidly adding good features and tools to an EMR. Some medical examples of open source modules include DrugRef (www.drugref.org) and our new decision support system (EGADSS, currently in design - come to the next THINK meeting for a presentation) that will be free to add into an EMR. In a closed source project, you have to pay for these licenses. For example, to leverage private vendors drug databases, such as First Data Bank, is prohibitive for most small private EMR vendors, leaving MDs in the lurch without any drug drug interaction support. There are also many examples of modules that can be incorporated from the non medical field to building an EMR (OSCAR, for example is built on the Tomcat server and MySQL database which are both open source).

c) Community - Beyond the license, successful open source projects expand to a community of people who embrace the idea of the project and are motivated to improve the project. You can certainly have people funded full time to work on an open source project, just like other projects, but you can get the benefit of additional help from a broader audience if you can encourage a community around you.

Linux has the benefit of having a larger proportion of users who have the ability to contribute code and there has been less of a “gap” between users and developers compared to EMRs, which makes it easier to communicate needs and translate those into useful software.. That is one of the reasons, I think, why Linux (and Apache) really took off as some of the shining stars of Open Source. It is one of the reasons medicine is so far behind on the technology curve in general (that and medical information is very challenging to model) – as you pointed out nearly all doctors are focused on doing what we’re trained to do – treat patients and there are such time and energy pressures on us, now more than ever. That is, in general, the challenge for us in this current reform movement as change takes energy and we don’t, as a group, have much to spare. But that is definitely another thread.

I don't think you need a huge number of core developers to be successful in an open source project, but you need a vibrant community full of excitement and a good mix of design, practicality, vision, and skills. Most importantly for an EMR, you need clear clinical input on what works and what doesn't and strong communication between the clinical and the IT members. But that goes for all clinical applications, doesn’t it?

You do need a critical mass and sufficient funding to see any vision through, open source expands that and the critical mass is not a big as we might imagine.

I’ll get off my soap box now. :)

Morgan

[Alan Brookstone]

Morgan, you make some excellent points. I look forward to your presentation on EGADSS. Do you know if open source EMR systems have been successful in other countries e.g. Denmark, New Zealand, US or the UK? Do you know of other examples in which sufficient adoption has taken place to create comprehensive implementations?

[isaac n silberman md`]

Combine open source electronic medical record with with real and scheduled time live interactive phone conferencing between patient and local doctor face to face and specialists elsewhere. In complex situations- phone/call center staff prescreening can prepare summaries for the phone conference. Unneeded tests, treatments and delays decreased most of time. We have worked out details of this model that can be applied .Details available for local adoption, control and implementation.

[Morgan Price]

Alan,

GNU/Linux And Open Source Medical Software News is a good place to start, as is the Open Source Working Group at AMIA if you are an AMIA member - IMIA Open Source Health Informatics Working Group

Morgan

[Jel Coward]

Hi Alan
I beleive VistA (Veterans Admin) is open-source (now) - take a look at
http://www.linuxmednews.com/

[peter richards]

one thought I have Morgan and all, that it would be good to have a rapid development prototyping system that a physician could mess about with and produce a sketchy idea of something interesting he wanted to do. Jel's idea of a template form for OSCAR f'rinstance. PHP might do it, maybe Eclipse or the new Sun Visual Tool but mysql/JSP technology is way beyond most people's ability -including me.

[Jel Coward]

For an in depth presentation of why OSS is a good way to proceed.

http://www.dwheeler.com/oss_fs_why.html

[Scot Mountain]

Unfortunately, I think open source EMR is destined to ride in the same back seat that other open source software has always taken with respect to commercial offerings (see Linux vs. Windows).

While the tech community has long lauded open source and predicted its eventual emergence, the truth is that, unless you are technologically inclined, open source doesn't offer anything that commercial systems can't outcompete. In other words, if you are one of the vast majority of software users, who is not prepared to code or otherwise adapt software, you might as well buy a pre-packaged commercial solution that is already set to go with the most popular defaults, as well as a phone number to call if something goes wrong.

While open source will always seem like a perfect idea to those that are tech-savvy enough to take advantage of it, most users simply lack the time/knowledge/ability/inclination to take advantage of it.

[peter richards]

but Scot - you can pay someone to install the open source application, train your staff and be on call for those disasters if you wish. and often at less cost than a commercial emr and you have the benefit of a community of users to get help from. with many eyes looking at your problem, it is more likely you will get assistance at least as quickly as a commercial vendor. you dont have to be a techno-nerd to use an emr. most users aren't.

[Alan Brookstone for Werner Keil]

(The following comment is sent in by Werner Keil regarding Open Source and Privacy relating to Personal Health Records - Ed)

Does anybody know a real working Eclipse RCP based EMR anywhere in the world?

I know, Central Europe got Elexis with growing user base in 2-3 countries (CH,AT,DE) but so far its focus has not been on English speaking countries or America.

Eclipse OHF has now been dissolved into something called OHT, but their projects so far contain nothing even barely like a Patient Management System or (end user friendly) EMR.

It looks like the US trends are "Cloud" focussed after all?! With systems like Google Health holding all patient's data and only RIA (Web) clients meant to access that. As opposed to most European countries where such models are still against their laws especially those protecting patient's rights and privacy.

Your thoughts on that?

[Jim Busser]

On the question of successful implementations, we might learn from the US Veterans' Administration's "VISTA" already offered above, even though it only became open source over time. To the best of my knowledge the VA has been associated with among the best (if not the very best) patient outcomes. If we stop to consider why, we might concede:

1. aligned with (and perhaps even driven by) the purpose of improving patient care. A commercial product is driven by the purpose of achieving revenue which, for EMRs, mean acquiring and *keeping* customers. There is a high threshold (barrier) among users to change EMRs as well as fragmentary communication among users and payors and consequently less leverage on the vendor to service and improve the EMR than might otherwise be the case.

2. cohesion on what it is that the VA's VISTA needs to do. I cannot imagine it is perfect, but the VA must have a better capacity than a disparate community EMR user base to achieve the cohesion. No doubt helped by what I understand to be collaborative discussion among clinical users within the VA on their patient outcomes.

3. control. The VA decided to control their own destiny. It is clear that a group of sufficient size could commission their own EMR... In ~ 2005 Newfoundland was close to identifying whatever open-source EMR best served their needs, and then putting their support into building it out, but the proposal died (or was killed). One might wonder at the role of fear-mongering by competition.

Accordingly I see three potential streams for the adoption of open-source. One is where the customer is of sufficient size and resoluteness to know and fund what it wants... the BC government took this approach for its EHR whose code it had commissioned and which will be operated by the coding contractor but it is BC that owns the code as well as the rights to get someone else to operate and maintain it.

At the other extreme is where open-source EMRs suitable for individuals or small group practices get built up and adopted over time... these EMRs do not have all bells and whistles... they need only adequately satisfy some not all of a practice's needs.

In the middle I suppose would be an EMR operated through some sort of consortium. This is being attempted in the US with OpenVista, where a government spin-off directs a process in which vendors provide support and added value... something like British Columbia's PITO but built on a single open source EMR (OpenVista). Variants include open-source EMRs controlled by commercial entities, e.g. OpenEMR, but I don't know how well such EMRs are doing.

[Jel Coward]

To answer Werner's question. No, I am not aware of an Eclipse RCP based EMR.

To come to open-source.
Wow! Reading the 2004 discussion above and the prediction (by some) that OSS EMR won't succeed and will take a back seat.

The year is 2009, just 5 years later. There is a FLOSS EMR (for Free/Libre/Open Source Software):

That has 509 practitioner ie. physician/midwives users across Canada (that can be counted, there will be more than that but because it is free, and download-able anonymously).

That is used for providing and co-ordinating medical care and administration for the City of Toronto's homeless shelters (via CAISI http://www.caisi.ca/wiki/index.php/Main_Page - the EMR has a complete shelter/client management system).

CAISI has won the Silver Award Winner in the Category of "Efficiency and Operational Improvements – Not For Profit" in the Canadian Information Productivity Award 2007
http://cipa.ca

OSCAR has a large user base in BC - over 100 clinical users - that is growing rapidly despite not being part of BC's part funded EMR program.

OSCAR has met all the compliance testing for OntarioMD and is growing rapidly in that Province.

The Departments of Family Practice at McMaster University and McGill University use this FLOSS EMR.

In other words - Open-Source EMR is showing its worth and its metal in Canada. With growth that happens with no advertising, just via the physician and academic communities.

.....and given the competition from highly govt funded, and highly advertised proprietary EMRs, this becomes something of a phenomenon to watch.

I am not sure where it fits with James' comments above. I think it is no more complex than a FLOSS EMR being successful because:

a) It works at least as well as proprietary offerings - so users want to use it.

b) FLOSS is attractive because it avoids 'vendor lock-in', data and physicians being held to ransom because they cannot escape from their single vendor supported, closed source product.

c) The existence of a vibrant community of users who can shape the software to meet their evolving needs.

d)The software/code being free to use and free to develop.

So the naysayers can do their bit. The proprietary vendors can spread their FUD (fear, uncertainty and doubt).

....but stopping a good thing is tricky, even with these tactics.

We have seen it all before - the Internet Explorer/Netscape/Firfox browsers wars - more recently the Windows/Ubuntu Linux desktop war. It will go on. Usually, the _right_ things prevail.