Wall Street Journal Talks About Open Source EMR and Vista

I’ve had a number of people ask me my thoughts on this Wall Street Journal article which talks about open source EHR and in particular the open source EHR developed by the VA hospitals called Vista.

I must admit that I’ve been enamored by the concept of free EMR. One of my most popular blog posts was this guest post about Free EMR by Medicare. Turns out that Vista is one of those open source (free) emr software that keeps popping up. I imagine it will continue to pop up for a long time to come.

Let me offer three points that I keep hearing over and over when I hear people talk about open source Vista.

1. (We’ll start with the good) Those that go to the VA are quite happy that no matter what VA hospital they go to, they have their information available. I’ve heard this on multiple occasions. I’m not sure if people are saying this because they’ve actually experienced it (which is likely considering the transient nature of veterans) or because they’ve had the concept drilled into their head. Either way, this is the major perception and considering it’s all one nice package I’m inclined to think it’s a huge advantage of Vista in the VA hospitals. I’d love to hear someone address how this “EHR interoperability” using Vista would work in commercial hospitals.

2. The users of Vista really don’t like using the program. It’s clunky, unwieldy and not the friend of the user. I’ve heard this multiple places and not just from doctors, but also from nurses and the IT people supporting the software.

3. The “database” that Vista uses, MUMPS, is a piece of junk and a major anchor on what could be an otherwise interesting open source project. I’m sure there’s some really interesting history behind the VA’s decision to use this MUMPS “database” system instead of one of the current SQL based database systems. Unfortunately, I’ve seen numerous people talking about the pains of MUMPS and the problem it creates for the future of open source EHR Vista.

I’ll admit that I’m not an expert on Vista, but I’m just telling you about the common themes I’ve read over and over again. Any other ones we should know about or other perspectives on Vista EHR?

About the author

John Lynn

John Lynn

John Lynn is the Founder of HealthcareScene.com, a network of leading Healthcare IT resources. The flagship blog, Healthcare IT Today, contains over 13,000 articles with over half of the articles written by John. These EMR and Healthcare IT related articles have been viewed over 20 million times.

John manages Healthcare IT Central, the leading career Health IT job board. He also organizes the first of its kind conference and community focused on healthcare marketing, Healthcare and IT Marketing Conference, and a healthcare IT conference, EXPO.health, focused on practical healthcare IT innovation. John is an advisor to multiple healthcare IT companies. John is highly involved in social media, and in addition to his blogs can be found on Twitter: @techguy.


  • John

    You have heard quite wrong and may want to do some more digging on VistA. This misinformation packaged as conventional wisdom does the taxpayer a disservice.

    In reverse order, MUMPS is not a piece of junk. It is the database of many of the commercial vendors and by far the most deployed database for clinical systems; e.g. Epic, Meditech, McKesson… It is a transaction oriented database that runs a hospital, and runs fast and reliably. It is an older technology but the longevity of its installed base is a testimony to its performance.

    VistA is very adoptable and usable. It must be as physicians who work at the VA are often rotating through for a month or two and must be able to get up an running quickly. Since the mid 1990’s, 65% of physicians trained in the US have rotated through a VA and have used VistA. The general consensus is that it is “straightforward” and saves them time. This can not be said for many commercial systems. Customers interested in VistA cite that their physicians are very comfortable with it. Its not been an objection what-so-ever, quite the opposite.

    Lastly, the market is becoming more and more focused on establishing health information exchanges. Quality care and alignment of patient care and providers within the system require electronic health records that are patient centered and support the continuum of care. The ARRA’S term “Meaningful Use” calls for incentives for the electronic sharing of records and the submission of quality reports. The HIE technologies are proven and the National Health Information Network has been demo’d. The NHIN connect software which allows VistA and other federal systems to connect is now in open source. The challenge of sharing records is no longer technical, rather it is one of organization and governance.

    Feel free to contact me with any further questions.

  • Edmund,
    I don’t think this information does the tax payer a disservice at all. I think it will cause them to ask better questions when evaluating Vista. These are common complaints I’ve heard over and over again.

    I find it interesting that you kind of make my case for me. The fact you list “Epic, Meditech, McKesson” as using MUMPS database is reason enough for people to avoid it. Plus, if MUMPS was so wonderful, then why haven’t other open source projects adopted it. I think you’ll find a hard time arguing that it’s better than MySQL or other open source databases.

    As far as usability, I’m just writing about things I’ve heard multiple times. I think in the article above it talks about a $7 million implementation of Vista compared to a $20 million+ implementation of some other commercial EHR software. Maybe it’s just hard for a lot of small office doctors that have a nice usable EHR for only hundreds a month to understand why a $7+ million project has software that isn’t as usable. Maybe it’s just comparing apples and oranges between ambulatory EHR and hospital EHR. I don’t think people have done a good job differentiating between the challenges of each.

  • I have some experience with MUMPS, which I viewed as “old” 15 years ago. I would have been right probably 10 years before that. By today’s standards, it is very slow, clunky, quirky, and very, very difficult to work with. MySQL or MS SQL Express would be superior solutions for almost any application. MUMPS has no true RDMS, as it isn’t really a true relational database.

    It predates true object oriented programming (some debate here) and has been around about as long as the first fax machine. By almost any metric, MUMPS would be very hard to defend as a technology suitable as a basis for a modern day EMR.

    I’m surprised that this is even a discussion. Here is one interesting article/post…there’s a lot out there if anyone is interested: http://thedailywtf.com/Articles/A_Case_of_the_MUMPS.aspx

    I can’t comment on Vista, but as far as MUMPS goes, I’m glad I got my shot.

  • Thanks for the link. This is the type of stuff I’ve seen over and over about MUMPS. I also forgot to mention that Dr. Billings above seems to work for MedSphere who I believe has built the backbone of their business on MUMPS with a product called OpenVista. Something I hear has been rather successful as an open source EHR project. However, I believe it’s an important part of his perspective.

  • MUMPS IS the underlying database for a lot of applications, and personally, I prefer standard SQL, thanks. It’s easier to maintain, easier to code for, and easier to take care of. MUMPS is unwieldy. Why don’t we all go back to programming COBOL? Because that would be silly. Why do we still use MUMPS? Because it’s legacy. You could cart the code around in a wheelbarrow, and I’ve found myself rewriting things more than simply maintaining code in ‘steady state.’

    VistA, from the experience I have had with it, has the same issues as any other EHR, only the database isn’t as easy to use as, say, MySQL, SQL Server or even Oracle. I guess my big pain point is that someone needs to do something with it — because it came out of my paycheck. I like the idea of open source EHR — I really want that to happen — but I’d really like to have a successful product that I paid for, and not something based on an out – of – date programming language that will require small doctor’s offices to hire folks who are probably running away from — and not towards — a MUMPS coding career.

  • Yes, I fully agree. The telegraph was pretty reliable too, but I just as soon use the telephone, instant messenger, email, text message, blog, twitter, myspace, or facebook to communicate these days.

  • Interesting – not one post so far from someone who actually uses MUMPS and VistA – so I’ll fill the gap. I started using MUMPS in 1982, and I’m using it today to continue development of a VistA implementation in a 1000 bed hospital. I’ve completed a 3-year Infomatics fellowship at CDC. I know relational DBMSs and SQL I’ll take MUMPS and its exceedingly flexible hierarchical database any day, and accomplish most jobs with 1/3 to 1/2 the lines of code needed by VB or JAVA.
    You need to read Phil Longman’s book, “Best Care Anywhere: Why VA Healthcare is Better than Yours”.

  • Actually, Jamie said that she had used it before also.

    Nice to hear that you’ve found some good in Mumps. No doubt there are reasons it was created and does certain things really well.

    I do think it’s important to point out that you’ve worked on it since 1982. That’s 27 years of experience using MUMPS. Of course you’ll love using MUMPS when that’s what you’ve done for 27 years. If you’d done relational databases that long you’d want to do it too.

    I know some people who love some of the things about Cobol. However, no one would use it today to build anything.

  • Realistic View about EMR/PHR/EHR
    Reference is made to this article….

    Making ‘Meaningful Use’ Well… Meaningful …Healthcare IT News web site
    by John Moore

    It puts some light on reality.

    It is a challenge to work with physicians – for both patients and vendors.

    Simple math is:

    You want to strengthen preventative care .. and or you want consumers to take charge of their health.

    Gov should try working more with consumers/patients than physicians or hospitals.

    Inovolving consumers with PHR and or giving them some type of incentive to take charge of their health can solve the situation better than comprehensive EMR.

    Why sohuld we use PHR:

    We, as patients, assume that we are not entitled to see or review our medical records. These documentsare usually viewed as being the sole property of doctors, hospitals or health plans. In fact, that is not the case, as these are your records and they belong specifically to you. Your medical record is a key tool for monitoring the care you receive. You should make sure that every detail of your records is accurate.

    One of the advantages of keeping your records onany PHR is you can check them for omissions. Notes from initial physical exams should cover your family history, lifestyle, past health, current health, and lists allergies and prescribed drugs.

    You will gain a great deal by taking charge of your medical records. Keeping your records on PHR is like keeping a journal of every interaction with your healthcare provider. It is time to give your medical profile the attention it deserves.

    If you notice any mistakes or inconsistencies in your records, change them right away. Send an e-mail to your doctor, explaining that the record does not coincide with your experience and ask why. Ask the individuals involved to adjust the medical records on PHR, and inform the physician, insurance company and/or hospital involved.

    If we can manage our bank account, we sure can manage our PHR

    Send your comments to White House Task Force

    Alexandra Govatsos Caulton, MBA
    PatientOS, Inc. Principal
    Speaking of meaningful use, I just recently flew home from a conference in D.C. and sat next to a pediatrician, we began exchanging information, and had a nice discussion on the hot topic of an EMR. He’s an unappy customer of Allscripts, and because I am not an expert in what that application can/can’t do, I can neither justify or discount his sentiments. However, he shed some light on the angst he feels when looking up a patient in his particular system. He feels that the admin work of doing a patient search, confirming it’s the correct visit/encounter to begin reviewing the lab results, charted data, and making clinical decisions is the particular piece of the workflow that stresses his physician staff the most. In the paper world, the staff pulls the correct chart for the physician, and the correct, most recent chart is on the top. Yes, there is some flipping of pages, but the analysis of confirming correct chart and encoutner is done for him. I tried to clearly articulate that even in the paper world there is no guarantee that it is indeed the correct patient, and in fact the likelihood of catching the error is worse than in the computer world. I site this example because it highlights the main issue of perception. Us clinical IT folks read this scenario and we can think of reasons how the process can improve, realize it’s not necessarily the fault of the system. Physicians encounter work flows like this in their own system, and feel that the system is not designed with a physician in mind.

    To your point on the PHR, I agree, and the patient portal is something all physicians, physician practices should implement along with their EMR!

  • The debate of MUMPS being old is strange. Who cares if it is 27 years old. Assembler is 50 years old and evne today all of the core operating system kernal is writen in assembler. You will be surprised that most of today’s Airline Reservations Systems are wriiten in assembler, I am taking about Sabber and Apollo. VisaNet transaction system is mostly writen is assembler. One needs to looks at the value of the product. It is fashionalble to throw out buz word and say it should use SQL. Do you know the Google and other decided not to use rational data base (SQL). Dose that make SQL bad?

  • I don’t think the main issue is that it’s old. It’s the way that it’s designed, the people to support it, the tools that interact with it, etc etc etc that are the core issues. You’re right that age isn’t as much of an issue as the other challenges associated with it.

  • Let us not forget the true nature of any EHR old , new or in between and that is the fact that human lives hang in the balance of the system’s ability to perform accurately, reliably, and securely. It would seem the VA’s VistA fits the bill. Let’s look at other critical programs like Air Traffic Control, NASA Space launch control, and Nuclear Missle Launch control aboard a submarine. My thoughts are that these systems are also old, but tried and true, and probably not driven by MySql.

  • hmm…..
    age old problem I guess – it’s not the age. it’s all about the capabilities, the flexibilities and most important of the useabilities for today and into the future.

    Having experienced mutliple HIS systems and especially MEDITECH (based on MUMPS) I can safely say that Relational Databases have much more depth and capabilities than transactional databases. One quick note on transactional databases, they have their benefits in certain application such as airline, banking where interoperability is not an issue!

    So back to the real issues with Healthcare today
    “Better Care” and “Reduced Costs”
    These politically charged buzz phrases can only be addressed by (to name the top few)
    – interoperbility so it removes all the redundancies
    – responsibility so the patients start to control their health a lot better
    – standardization – this is a biggy and will be forever debateable but let’s start with the best one available today and finetune along
    – legalities. It will be great if the Lawmakers come up with policies that govern privacy so patients online records are better protected and thwarts anyone other than the ones vested in the best interest of that particular patient.

    EMR/EHR/PHR has got to happen. It’s all about the Patient and the Patient’s health. It’s priceless.

    I would love to be a part of any forum or entity that addresses these issues. Count on me

    Kudos to all who voice their concern and their perspective – even if it’s negative or non supportive!

  • Seems like a lot of effort, running down a system you haven’t used or don’t understand completely.

    The fact is the Vista is built on Mumps or M programming language – which IS the database, the code, the software – much like a DB2 or Oracle design might be built. On top of that base there have been built GUI interfaces to ease the uptake by physicians and nurses. These interoperate to form the most successfully adopted EHR in the world today, and to improve patient outcomes in the VA healthcare system.

    There have been tens of thousands of hours invested by healthcare providers, support personnel, programmers, and administrators into the Vista product to acheive successful EHR operation and to improve the care provided to veterans. This time is what the taxpayers have invested in, and the result is a system of long-standing success that overshadows the EHR world today. The fact that development has been ongoing for almost 30 years is a testament to the commitment of a team of individuals and of an institution to the success of a project.

    And to keep the playing field level, please consider that most legacy systems tend to remain on their legacy platforms until those platforms are no longer viable – be it hardware or software platforms in question. The hardware Vista was developed on, Alpha, is no longer a viable option; however the Intel Itanium chipset has taken up where that left off, and HP is supporting the hardware going forward with the same OS (OpenVMS). Additionally to scrap thirty years of development is a terrible tragedy, and while there are projects underway to port M to other languages, it is the business rules built into the code and into the system of a package like Vista that make that challenge huge and unwieldy. Those same rules, actions, and triggers are the things that insure patient care and provide “checks” on the providers practice of care and medicine.

    Overarching all of this is the fact the Vista is much more than and electronic health record, and it all but an ERP for hospitals – the modules (I think 180+ last estimate I heard) are many and varied in use and needs. Vista is HUGE and provides much more to any type of care establishment (from single provider practice to multi-hospital organizations) than simple CPOE and vitals recording. This in itself would present a problem for redevelopment or porting of the code. Rather than scrap the dollars spent by the VA over the past 30 years, the path forward here seems to be to build upon those successes and extend the care model provided by the Vista software into healthcare institutions and doctors offices nationwide. Vista keeps on running, churning out better successful outcomes and demonstrating the need for EHR to aid physicians in providing solid records and continuity of care to patients. And the M developers are out there, in droves. They believe in Mumps as much as anyone else believes in the language they were taught and rely on to do their work, and I don’t believe the answer is to toss it out or write it off, but to look for ways to integrate with it and with Vista and find the synergies with new platforms and add capabilities to the system. (Mechanics still use their old wrenches to tighten bolts don’t they? If a new bolt get’s made – they’ll buy new wrenches too).

  • JD,
    I think it’s very clear that there are a lot of people that don’t like MUMPS/Vista and there are people that have worked on it and see its benefits.

    However, I think it’s really funny that you think that Vista will work very well in a single provider office. I’d love to talk to someone who’s implemented Vista in a single provider office.

    One other thing that I think you’re missing is that the VA is a really unique and integrated environment. It’s not very representative of things like the multiple insurance companies that most hospitals have to deal with. Of course, that’s just one example.

Click here to post a comment