Network News

X My Profile
View More Activity

Open-Source Health Care

So long as I'm talking about hidden dangers in health-care reform, it's really worth giving Phil Longman's article on Health IT a read. I think he's a bit too aggressive in advocating for a particular solution (the Vista platform, which I like, but which has its problems), but he's really right that the final product should be open source. This is one of those places where a bit of geek theory could really do a lot of good.

By Ezra Klein  |  July 14, 2009; 5:35 PM ET
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: Why Health-Care Reform Begins in 2013
Next: A Win For Medicaid


Your comment about "Vista" will surely confuse many readers (especially those who don't read Longman's article in Wash. Mo.). First, VistA is the name of the open source medical software developed principally by the Veterans Admin. "Vista" is the name of a Microsoft operating system, which surely is a platform, but is surely not medical record keeping application software.

Longman's article mentions somewhat favorably a proprietary health IT product from Epic. Since Oregon Health and Sciences University's hospitals/clinics use Epic, I have a patient's eye view of that product. It it certainly comprehensive. It is mind-boggling complicated as well. I've seen several OHSU physicians struggle over the learning curve and into regular use, and it is very cumbersome for the physician to use, and even the web-based patient remote-access software (to their own medical records) is no better than 1997-99 era web applications (it sucks!).

If Epic is the best of the proprietary software for health IT, then none should be considered. Here's a small example: Epic has no way to flag a key health message about a patient. I have a metal prosthesis in my left middle ear, making me ineligble for MRI imaging. The metal would turn into a bullet to the brain. Yet an ER medic/nurse would not learn that unless they scanned literally dozens of screens in Epic before doing anything. There are many conditions like mine that are first-order imperatives for medical providers. I'm reminded of an episode of House (MD) on TV where House himself needs surgery on his right leg. House writes on his left leg: "NOT THIS LEG".

Lobbyists can do evil things. This health IT thing is way too important to leave up to lobbyists and congressionals to decide.

I've wondered myself how to manage a medical emergency involving myself - particularly where I was not conscious. A 911 ambulance will take you where they want - which is very probably not to OHSU. Yet all my medical records are locked up in Epic at OHSU and would be unusable at any other hospital, including those with Epic software.

[my comment will be continued in another post].

Posted by: JimPortlandOR | July 14, 2009 6:26 PM | Report abuse

Final comment: In the 1970's I worked for Hewlett-Packard Company's then-new computer division in Cupertino CA, in marketing, and later in R&D.

I was sent to visit several HP customers or potential customers in Europe one summer. One of the countries I visited was Sweden, where we were taken to the Karolinska Institutet (Karolinska Institute) which is one of Europe's largest medical universities. It is also Sweden´s largest centre for medical training and research. This was about 1973 or so. They had computerized records for every single person in Sweden online and available to all physicians in Sweden. One system, one set of screens for all providers to learn. Forty-five-some years ago!

Yes, Sweden is smaller, and today a single mainframe system would not be the right technical solution, even in Sweden (they likely have evolved as well). But the key thing was standardization of the medical record and software displays to update and retrieve the information. We should DEMAND no less.

Posted by: JimPortlandOR | July 14, 2009 6:38 PM | Report abuse

yeah, but the EHR's aren't going to be from any single maker anyway. They'll have to meet format requirements, testing requirements, and probably some certification requirements tied to testing, but providers will be able to use any of several options. They won't be tied to one winner-take-all. Longman's article reads a little bit like he thinks the government is going to select a single vendor and he wants that vendor to be open source. He uses sort of shoddy logic. HIMSS cares that Stark's legislation was going to put their members out of business, HIMSS - it is safe to assume - was significantly less concerned about the 'manner' in which the legislation would've put their members out of business.

I would, however, be curious to see how the open source products fare against the closed source products on the open market. I'm sort of ok with letting the market settle this fight.

Posted by: ThomasEN | July 14, 2009 6:40 PM | Report abuse

I don't really see why Healthcare IT platforms ought to be open source. I think open source platforms ought to be considered, and if they are favorable in terms of functionality and total cost of ownership (OSS is only free if you can get IT professionals to work for nothing).

In general it's not really the case that open source software is better than the proprietary alternatives. There are a couple of OS standards (Linux, Apache, Firefox) that have succeeded by having large dedicated communities but you'll notice that the OSS products that are continually maintained and improved are primarily tools for developers themselves.

Reading Longman's article, it sounds like VistA has an active developer community, and an organization (the VA and DoD) that are committed to maintaining it.

However, it being open source isn't what's salient about VistA, it's the fact that it's a good piece of software that does what it's users need it to do. There's no reason that couldn't come from a proprietary application as well.

Posted by: TWAndrews | July 14, 2009 11:07 PM | Report abuse

"OSS is only free if you can get IT professionals to work for nothing"

Of course, commercial software is not free even if you can get IT professionals to work for nothing, and there are fewer IT professionals who are experts in any given commercial solution, or who enjoy working with them.

Posted by: albamus | July 15, 2009 5:30 AM | Report abuse

Oh, it absolutely should be open source. We use proprietary software that seriously takes a year or two to fig a bug and three to five years to develop and implement new features.

Nothing kills innovation faster than, "sorry. The computer doesn't work that way."

Posted by: Aatos | July 15, 2009 8:57 AM | Report abuse

This is totally ridiculous. Let me see if I can put the standard open source argument in terms everyone can understand:

1. Require computerized health records
2. ???
3. Open Source!

If you're focusing on the code, you're completely missing the point. The important thing to get right is the *data*. How are my records stored? How are they protected? How are they shared? Code is transitory. Data is *forever*. Once you've figured out how we encrypt the stuff going over the wire, what format records need to be in, etc., nobody should give a crap whether it's accessed from Windows, Linux, or the Mac OS, from a browser, a standalone GUI application or a command-line script, open or closed source. The software matters if you're worried that you won't be able to get a fix for a security breach, but there's no evidence open source is any better than anything else. It would also matter if you suspected that the software were malicious, like for say voting machines, but I don't buy any argument that there are any similar incentives to mischief with medical software.

Posted by: BlakeTheGeek | July 15, 2009 9:14 AM | Report abuse

Open source doesn't have a lobbyist.

Do the math.

Posted by: pj_camp | July 15, 2009 9:35 AM | Report abuse

I agree that the software should be open-source. However, Longman's emphasis is in the wrong place. The single most important criteria is that the standard itself should be open.

To draw up an example, RSS is an open standard. I can read RSS feeds using (to name a few examples):
Microsoft Outlook (closed source software with hefty licensing costs)
Google Reader (closed source but free to use)
Thunderbird (open source and free to use, modify, distribute the code)

Honestly, the people publishing blogs and newspapers don't care what software is being used at the other end - the important thing is that the entire suite of internet standards allow for computers on the internet to play well with each other regardless of how much the underlying software costs or how liberal the software license is.

It's nice to be able to send someone an email without having to first ask "hey do you use Mac, Windows, or Unix/Linux?"

In the same way, health IT should first and foremost be done with open standards - publish the specs of how computer systems will talk to each other, how database systems will synchronize health data, etc. It doesn't matter if a for-profit closed source software solution comes along, as long as it doesn't lock hospitals/practitioners/patients into one vendor.

To accomplish this, I propose that we establish an organization structured like W3C or IEEE to define and refine the standards, inviting voices from every interested party - software/pharmaceutical/insurance corporations, hospitals/clinics, doctors/nurses, and patients.

Posted by: shanehuang | July 15, 2009 12:01 PM | Report abuse


The standards are open, and already in use:


Posted by: ThomasEN | July 15, 2009 12:47 PM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2010 The Washington Post Company