Open Source as a Healthcare Solution

The headlines in the US media warble every day with dispatches of how the government wrestles with the question of providing better healthcare for the nation’s residents. The debate now in Congress has proven to be a lengthy one, and there is little doubt it will be contentious right up to the passage of any legislation (and, most likely, beyond).

But as the US watches this debate unfold, many constituents may not realize that steps have already been taken to improve the US healthcare system, both with recent legislation and with 30-year-old software code–code that one company is harnessing with open source practices to improve clinical care across the country.

The code is from the VistA project, an effort started by what was then called the National Center for Health Services Research and Development of the US Public Health Service to create an electronic health record system in the 1970s.

The system is not an application, but rather an architecture for different modules that work together to track the records of a given patient. It was implemented in the Veterans Administration (VA) hospital system, and as time went on and more medical practice and experience was integrated into the system, the architecture that would be known as VistA would become so efficient, it has received more than a little credit for helping the VA healthcare system, Veterans Health Affairs, turn itself around in recent years.

The interesting thing about VistA is that the entire architecture is public domain, which is even more free than the GPL. That means anyone can take the code and use it however they want. So anyone who wants to build an electronic patient record system has an $8.6 billion headstart investment in the code, courtesy of the US taxpayers. This has led to the deployment of VistA systems in several non-VA (and even non-US) hospitals, and quite a strong industry of firms that have grown up to implement VistA tools.

A leader in this space is Medsphere, which manages OpenVista, an system that in very simple terms is VistA with a GTK/C#-built front-end. Even better, it can run on Linux.

“Manages” is actually a pretty good word for what Medsphere does. The OpenVista code is worked on by a growing community of developers, which include “hardhats” (ex-VA coders who relish digging deep into the guts of the VistA code) and younger developers whose skillsets lean towards web interface and application development.

Medsphere works on OpenVista as a community participant, too, providing its own modules that work in OpenVista. It also acts as project mediator, prioritizing the direction that OpenVista (and to some extent, VistA) will go.

This community-based approach is nothing new to even the most experienced VistA programmers. According to Rick Jung, Chief Operating Officer of Medsphere, “the community was originally a skunkworks for VistA development.” That built-in community carried over from the public domain development into the free software development ecosystem that Medsphere harnesses today under OpenVista’s Affero GPL.

Medsphere is sitting at what one could call, with little hyperbole, a crossroads in history. Not only is the larger healthcare debate going on, but there already is an healthcare stimulus in place that US medical facilities are working towards.

Known as the Health Information Technology and Clinical Health Act, this law was passed early in the Obama administration’s term as part of the larger American Recovery and Reinvestment Act. Essentially, the stimulus will provide $19 billion over a four-year period to facilities that adopt and use electronic health record systems in meaningful ways.

(As a side debate, healthcare IT workers have spent a lot of time figuring out just what defines “meaningful use.” Currently, the presumption is 10 percent of healthcare records entered into an e-record system.)

Whatever the threshold, a $19 billion stimulus pot is a big incentive for hospitals to get their systems shifted to a e-recordkeeping. Once the meaningful use standard is met, they can use the stimulus money to finalize the conversion.

This leaves Medsphere looking at a much larger immediate customer base. Even better, according to Jung, is the fact that around 65-75 percent of practicing physicians have trained at a VA facility. Which means, Jung explained, they’re familiar with the VistA system already.

That’s a pretty sweet market position, and one hopefully Medsphere can capitalize on. Actually, based on my conversation with Jung, the benefits of streamlining the recordkeeping process are very desirable no matter who’s providing the technology.

Here’s an example: when OpenVista is deployed in a mid-sized hospital like Midland Memorial Hospital in Midland, TX, the customer saved about $6 million in licensing fees they would have paid a proprietary software firm. (Medsphere uses a subscription-based pricing model.) But the real savings, Jung explained, comes in the prevention or lessening of complications that deviate from best-case scenarios.

Say you go to the hospital to have a benign brain tumor removed. While not a simple surgery, the basic flow of care is pretty predictable in the best case: go in, get some good medication, get a hole in your head, have tumor removed, patch hole in head, have tumor tested (just in case), wait until you can go home, then go home.

Now that we’ve all seen why I did not go to medical school, let’s presume that in this best-case scenario, the patient is in and out of the hospital in a week, with no complications. For the patient, this is good, and so, too, for the hospital. All of their resources were used efficiently for the task at hand.

Let’s assume things didn’t go so well. Despite excellent care, complications arose, such as the patient becoming septic (which is an infection by some nasty bacteria that, if not managed well, can lead to organ failure or death). Now the hospital is diverting more and more resources to the patient (whom it still must treat), who is fighting a harder battle toward recovery and racking up massive costs for the patient, their insurance company, and the hospital.

It should be noted that healthcare institutions plan for these kinds of things. Despite the best of precautions and preventative procedures, complications will happen. A reality-based best case, Jung explained, is not when everything goes perfectly, but when a clinical facility has “the ability to troubleshoot, enabled by technology, so troubleshooting early will minimize complications.”

That’s where OpenVista’s open development methodology can be a big help. Healthcare workers who use the OpenVista system in various locations participate in the software user groups, Jung said. Recently, a group of user-nurses shared information that, after a patient’s vital signs were entered, certain symptoms that showed up might be a sign of sepsis, such as a sudden drop in blood pressure.

OpenVista developers noted the discussion, got more details from the nurses and other clinicians, and now the software has a feature built in that will flag users if certain danger signs are met. The patient may not be septic, but if they are, the earlier it’s discovered, the more the potential damage can be mitigated.

This is just one example of how open systems can be more efficiently improved. And these kinds of improvements don’t just make software easier to use, they also save lives.