Issues with VA Vista EMR

So many people have propped up the VA’s EMR system (Vista) as the model for how EMR should be done.  This story about the GAO finding the EMR implementation over budget is really interesting.  Here’s just one short section about the budget that they have for the VA EMR:

VA officials cited resource availability and interdependencies among projects as key drivers of cost and schedule variances. The GAO has estimated that the program will overrun its current budget – worth approximately $1.897 billion – by $350.2 million.

WOW! That’s a lot of money. I would hope that if you’re spending close to $2 billion you’d have something good to show for it. Too bad it’s just not reasonable for most doctors offices to spend that kind of money.

Here’s another interesting quote from the article (emphasis added):

VistA-FM is designed to provide a framework as well as additional standardization and common services components. It’s also intended to eliminate redundancies in coding and support interoperability among applications. However, VA officials have told the GAO that VistA-FM is costly and difficult to maintain and doesn’t integrate well with newer software packages.

I’m sure the MUMPS fans will come out of the wood work and tell us how great it is. I’m sure it does some things very well. However, I agree with the quote from this article is that it doesn’t integrate well with newer software packages. This is a major problem if we’re talking about inter operable EMR software.

Vista is free for doctors offices. I think it’s the “difficult to maintain” issue that kills most people even with the free price tag. Of course, my focus is on ambulatory EMR. The hospital environment is a mess regardless of which EMR you choose.

About the author

John Lynn

John Lynn

John Lynn is the Founder of the, 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,, 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.


  • I read your comments with interest and would like to add a few more thoughts about VA Vista’s success and the state of the US EMR marketplace. Perhaps the most important aspect of VA Vista’s success is that it has cultivated reasonable expectations and been backed by strong policies and procedures to keep it reliable and trustworthy within the limits of its functionalities. This is no small feat and it mirrors in many ways the significiant successes of the the DoD’s cousin to VA Vista, CHCS I (not to be confused with AHLTA). VA Vista is completely “interoperable” across the vast VA enterprise, also no mean feat.

    In contrast, in the civilian marketplace, US EMRs remain non-standardized, non-interoperable, and to the extent there is a certification process, it is currently not addressing basic and long-established legal records compliance requirements. Given the policy decision to implement non-standardized and non-compliant (in some cases defective when assessed as records management systems) and fix the problems later, we will have an extended period of time where the early adopters are stuck with legacy (obsolete) software dating from the current non-standardized, non-interoperable, minimally certified marketplace. Worse, the stakeholders benefiting from this “implement now, fix later” have little or no incentive to highlight this “buyer beware” marketplace and equip EMR purchasers with tools to provide means to separate the good from the bad from the ugly.

    This legacy of untrustworthy primary source systems will have pervasive impact since all secondary and derivatives uses of information from systems that themselves are highly variable (and, in some extremes, of overtly defective design) must necessarily be at risk of significant variability and so untrustworthy. Any statements about their necessary utility in all instances for improving health care in their current state are therefore false. This is the fundamental problem with HITECH’s stimuli to uptake, that EMRs will be uncritically taken up as if they were in fact standardized, interoperable, and meaningfully certified at this time.

    This underlines the importance of the fact that HITECH processes are also poised to (potentially) markedly improve the meaningfulness of EMR certification. Those near-term decisions will have substantial impact on how much trouble HITECH ends up creating for bedside record trustworthiness, as well as how long that trouble will last. Maybe, with some new improved compliance-oriented certification requirements and use testing the term “shovel ready” will become modified into a humorous euphemism about how many EMRs will need to be buried deep underground to make way for truly meaningfully trustworthy and useful ones?

    The good news however is that where there is careful, methodical implementation of even today’s EMRs, configured properly and addressing verifiable problems in line with basic principles of information integrity, substantial results are achieved (see, for example, HIMSS Davies awards and their winners).

    I hope this site will take up the cause of improving the transparency of the EMR marketplace with regard to clinician/user/patient trustworthiness assurance and risk reductions would be very welcome. I have no doubt that the US is on the way to eventually achieving reliable, trustworthy EMRs at the patient’s bedside. This is a goal that reasonably both precedes and enhances all other derivative uses for EMR-sourced information.

    VA Vista’s successes, though not 100%, merit further illumination to better inform next steps. Highlighting what DOES work about it for end-users and for patients, will be much more helpful than focusing on its warts. BTW, if you think $2B is a lot for an EMR, look into what the DoD has invested (of our taxpayer dollars) in AHLTA.

    In the meantime, here are some publicly available resources to help inform EHR selection and use.

    A simple, basic EHR testing and scoring tool:
    Gelzer, RD, Trites, PA “EHR Exam: Using Test Vignettes to Assess EHR Capabilities,” in Journal of AHIMA, Vol. 77, no. 5 (May 2005): pp. 56-59.

    Migrating Forms Processes from paper to the EHR:
    Westhafer, K., “The Forms Management Process: Keeping Pace with EHR Development,” Journal of AHIMA, Vol. 76, No. 8 (September 2005): pp.66-67.

    AHIMA e-HIM Work Group on Maintaining the Legal EHR. “Update: Maintaining a Legally Sound Health Record—Paper and Electronic.” Journal of AHIMA 76, no.10 (November-December 2005): 64A-L.

    AHIMA e-HIM Work Group: Guidelines for EHR Documentation Practice. “Guidelines for EHR Documentation to Prevent Fraud.” Journal of AHIMA 78, no.1 (January 2007): 65-68.

    AHIMA Copy Functionality Toolkit

    A more complete testing regime is available for purchase thru AHIMA’s bookstore:

    How to Evaluate Electronic Health Record Systems

    By Patricia A. Trites, MPA, CHBC, CPC, EMS, CHCC, CHCO, CHP, CMP(H) and Reed D. Gelzer, MD, MPH, CHCC AHIMA (2008)

  • Reed,
    You make some good points about the problem of there not being any standard for interoperability in the EMR industry. I think that this site and my sister site have done a lot to educate doctors on how to have a successful EMR implementation. I think this is the first step in the right direction.

    I guess for me the problem isn’t a technical one at all. Even if you implemented Vista in every doctor’s office, we’d still not be any closer to EMR interoperability. The technology is the simple part. The reason the VA can talk amongst all of its locations is because it owns and manages all of them. This is the same for someone like a Kaiser.

    So, until we solve the legal, governance and funding issues for EMR interoperability we don’t need to worry about the tech. If those other things were solved EMR software would be interoperable over night.

  • John,

    You jumped at the wrong conclusion when you jumped on VistA as being the faulty item here. What has failed is the “-FM” portion of the GAO report, the Foundation Modernization. You see, VistA is NOT VistA-FM. VistA-FM is the effort to dismantel VistA. Just like all of the other Attempts in the past nearly 20 years, these efforts are under-functioned, over-priced, and way over their delivery schedule. A mere fraction of the cost of what has been expended to replace VistA would have made VistA able to totally out-class every other approach to EHRs. There is work currently going on in the Open Source community to extend VistA and it is working very well. Here are some of the projects that are currently on the way or already in production;

    Lab, while the VA is outsourceing to Cerner (with interesting results), the rest of the community outside the VA is continuing on with enhancements and options that will make it easier to install and higher functioning as well as affordable to nearly everyone.

    Continuity of Care Records and Data (CCR/CCD) while this standard is a bit anemic, it does promise that we might be able to project all of the VistA databases to other systems or accession data from others.

    Holographic EHR – This is one of our concepts, basically you could think of it as “VistA for One” (or a small group of patients), a self consistent subset of the parent VistA environment which could be booted separately. The self-consistent “VistA for One” becomes a mechanism for complete transfer of patient data from one site to another with merge capability. It also becomes an in-hand user copy of his records which can be protected via a network keying system which registers the data set, and records the efforts to open the data set and by whom, and who is attempting to accession the data to what target VistA system.

    This is fun. I cannot tell you the number of times that I have heard, we need to keep CPRS, but get rid of VistA. The engine behind CPRS IS VistA. Without VistA, CPRS is a screen-saver. The Open Source Community is making enhancements for the CPRS/VistA environments. There is another group that is working on the webification of VistA with open source tools.

    By the way, I worked on the proposal team for CHCS-I and we used MUMPS to build interfaces for various other vendors to communicate with each other. In fact, the MUMPS interfaces worked better than the Clover-leaf connection engines.

    There is a reason that the Subject Matter Expert developed systems of the VA, DoD, and IHS have been so effective and difficult to replace. VistA is a whole enterprise solution that the vendors hope you never find out about. The vendors focus on dismantling VistA to provide a new niche to build “customer loyalty” (make it too painful and expensive to move to something else so the customer is essentually stuck with the vendor’s solution only. With the VistA model the SMEs are the folks at the point of care, and not a programmer who has never spent an hour in a hospital, yet is charged with the setting of policy for the hospital in his interpretation of the requirements (which may or may not reflect the intent of the SMEs).

    By having VistA as Open Source, this means that the cost of doing development has dropped right into the basement. Success can be tried in a thousand places, but with Open Source, as soon as someone comes up with an enhancement or corrects a problem, the change can go out to the rest of the World. The best of breed solutions float to the top to be applied everywhere.

    You know, VistA is still running the VA hosptials for over 30 years, don’t you think that if the vendors could have replaced it, they would have? They have tried and gotten paid well for the attempts. But this is part of the problem. There is no incentive to ever complete a task or attempt because then the paydays end. This is why they have confused the community with the use of VistA-FM, use their failures as justification to try to replace VistA yet again.

    Let’s take a look at some of these magnificent failures. How about the replacement of IFCAP (the financial part of VistA) with Core-FLS. Now get this. The VA developed IFCAP (by the way, it was not vendors who did this work, it was the VA SMEs who did the daily work of inventory and supply and finance) and owned the code. The VA paid nothing for the code other than the VA programmers and SME’s time. Then they were going to replace it with a package which would only have to do 30% of what IFCAP did. Congress committed $470 million to replace something the VA already owned with something that had less functionality but was more glossy and the VA would have to pay big bucks to the vendor to support. The roll-out of the product was done at Bay Pines VA Medical Center and was so bad that they had to close elective surgery. The vendor spent over half the money just to install the first site and the project was mercifully stopped and IFCAP was re-installed. So much for modernization. This is not an isolated incident.

    There was the Spanish Pharmacy labels. Peurto Rico and many of the boarder VA Medical Centers needed to be able to produce Spanish Labels for the Hispanic Patients. This was done by duplicating code rather than completing Internationalization that was started back in the early 1990’s, but stopped by the Clinger-Cohen Act. It would have taken less time and less money to complete internationalization for all of VistA than it took to do a one-up parallel code base for Spanish Pharmacy Labels. Adding another language would mean even more complexity (such as French or German), would be even more duplicate code for a single functionality. By myself, I built a tool to convert all of VistA into being ready for Internationalization and made it so there could be any number of languages that could be selected by the user and not necessarily locked to a single language. It takes about 50 minutes to parse all of VistA into the instrumented code and load the DIALOG file with the words and phrases, ~165,000 phrases in all on a 800 mhz laptop. It does not modify the distributed code but builds the instrumented code in a separate location. This code is available for free download from WorldVistA.

    The community is alive an well, and vibrant with new ideas. We are starting to catch up from the “legacy era” and allowing the evolution of the tools to progress again. Want to join in?? It is a lot of fun and a set of real challenges that will bring the power of what needs to be done, back into the hands of the people who are at the point of care. Interesting thing about the word “Legacy”, people think of it as old or non-functional. It really isn’t. It also means that the code is doing the job and doing it just fine. Can it be improved, sure, VistA was made to be improved, to expand beyond what was known and what was learned. But, do remember, VistA-FM is NOT VistA, it is the attempt to break up the integrated hospital system into a series of stove-pipes. VistA-FM is the worst of all FUD (Fear, Uncertainty, and Distrust). VistA is still running the hospitals and it is running more community hospitals every year.

  • Sorry to hear about your problem with the disease, mumps. Be assured it has nothing to do with the computer language which was the 3rd ANSI Standard Language after FORTRAN and COBOL. MUMPS, the language is still very much alive and well and running most of the successful hospitals in the world. It is very useful there and in Banking and Credit Unions in the world. It is still being used in the travel industry as well as in a lot of legal firms for putting together legal documents. It is still being used to compile the Yellow Page Phone Directories as well as other applications most people would be very surprised to hear about.

Click here to post a comment