How Does the VA’s Technology Rate Against Other EMR Vendors?

Health care for veterans has been all over the news.  At the same time, the DoD is moving to procure a replacement EHR system.  So it seems there is no time like the present to review a recent RAND case studies report entitled “Redirecting Innovation in U.S. Health Care: Options to Decrease Spending and Increase Value.”

The case studies include a chapter comparing America’s two most broadly deployed EHRs:  The VA’s VistA and Epic.  The tale RAND tells is not one of different EHR technologies, as both VistA and Epic both employ the MUMPS programming language and file-based database. Rather, it is about how different origins, business models and practices have dramatically influenced the respective systems.  As the report itself says, the contrast offers “useful insights into the development, diffusion, and potential future of EHRs.”


VistA, “the archetype of an enterprise-wide EHR solution,” supports the Veterans Health Administration, “the largest integrated delivery system in the United States.” Initial VistA development was a collaborative, distributed, grass-roots effort where individual VA medical centers built out new clinical functionality on a common platform.

In the mid 90’s, VistA became the instrument of change at the VA.

The pace and scope of EHR adoption increased dramatically under the leadership of Dr. Kenneth W. Kizer, who served as the VA’s Undersecretary for Health from 1994 through 1999.  Dr. Kizer considered installation of a major system upgrade to be a core element in his effort to transform the organization …

Full Disclosure: Dr. Kizer is Medsphere’s board chairman.

From February 1997 to December 1999, the VA connected VistA across all agency facilities. “It was (and remains) the largest and most rapid deployment of an EHR ever done,” the RAND report says.  As a result, from 1996 to 2004 the VHA realized dramatic improvements.

  • 70 percent increase in patients accommodated
  • 37 percent decrease in employee-to-patient ratio
  • 20 percent decrease in overall costs per patient (Medicare up ~20 percent over same period)

Among other factors, the VA’s success has been attributed to the collaborative development of VistA, as well as its clinical adoptability and interoperable data sharing. RAND notes that VistA’s clinical acceptance has been consistently validated over numerous surveys:

Independent surveys suggest that physicians are broadly satisfied with VistA compared with other EHR systems. In a 2011 survey by the American Academy of Family Physicians (AAFP) and a similar 2012 poll by Medscape, VistA outscored the large majority of health IT solutions, particularly those offered by large vendors like Epic and McKesson. When the AAFP survey asked respondents to express their level of agreement with the statement, “This EHR enables me to practice higher quality medicine than I could with paper charts,” VistA received the top score. In both the AAFP and Medscape survey’s, VistA was one of the highest ranking systems overall.

As public domain software financed by the American taxpayer, VistA is affordable for any hospital or health system and it clearly demonstrates the promise of EHRs in meeting the goals of the HITECH act: adoption, acceptance, Meaningful Use and interoperability.


The RAND report looks critically at Epic’s dominance, which few can refute at this point in the history of EHRs.

Epic has established itself as the enterprise-wide solution of choice for large private health care systems and academic medical centers, irrespective of ongoing concerns about its limited interoperability and less-than-ideal usability.

There seems to be a natural history to all technology dominance.  There was a time when Microsoft, for example, was the most valuable company in the world.  I think Epic’s closed-platform approach has a similar trajectory and, similarly, over time I believe the company’s proprietary single-source model will erode its dominance. This is what happens when you put up walls between your end users and innovation / interoperability.

An independent evaluation of Epic’s impact in the Kaiser system found that implementation led to efficiency losses and a persistent two-minute increase in the length of time of an average patient encounter … Moreover, physicians complained of workflow interruptions, slower processes of care, and excessive time with the provider’s “back to the patient” because of the need to focus on computerized order entry.

While Epic preaches interoperability, it practices non-interoperability and vendor lock.

One of the biggest concerns with Epic is its relative lack of interoperability. Although the company has a strict structure and retains tight control of its software and data, it does customized installations for each client … it limits Epic’s interoperability between sites and much of its capacity to communicate with out-of-network providers who use other EHRs. In some health care systems, two facilities running Epic in the same health care system cannot share data.

Quoting Epic critics, the RAND report goes on to say that perhaps Epic’s “Care Everywhere” interoperability solution is disingenuously named.

EpicCare (Care Everywhere) specifically breaks the standard CCD form, and makes it incompatible with the rest of the 400+ EMRs in the USA by adding their proprietary extensions. This is consistent with Epic’s proprietary, one-vendor-shop, non-interoperability stance. The statement that “any hospital can interoperate with Epic’s Care Everywhere—just so long as they are an Epic institution aptly summarizes this.

Epic’s weaknesses are VistA’s strengths:  affordability, usability and interoperability.  The Affordable Care Act and its intended improvements in quality and cost depend on a network effect driven by these three characteristics; to date, the dominant business models have trumped the promise of innovation and interoperability, which RAND believes is both the key to making EHRs worth using and a threat to the status quo.

The shift (to greater interoperability and more openness) will be less welcome to large legacy vendors because it will blur the competitive edge they currently enjoy.  Health care systems may be less-than-enthusiastic adopters because functional HIEs will make it easier for patient (sic) to see nonaffiliated healthcare providers or switch to a competing health care system.

Moving forward, both systems will continue to have success, but only one can support the innovation and interoperability necessary to transform U.S. health care. As I understand it, this was the original intent of the ACA and HITECH, not ensuring a ready market for health care related businesses.

– See more at: http://blog.medsphere.com/vista-and-epic-a-tale-of-two-systems/#sthash.M26uIgve.57YwaJJq.dpuf

Categories: Uncategorized

Tagged as: , , , , ,

3 replies »

  1. I would argue that the system the VA used PRIOR to the rollout of VISTA was actually better.

    I trained in the mid 90’s and spent massive amounts of my time at a huge VA center, then, as an academic physician, was also associated with another (different) large VA hospital.

    In the early 1990’s, the VA used paper charts, but all labs, all x-rays, all H&P’s, operative notes, and discharge summaries were on the computer. We used rather “stupid” desktops and had monochrome monitors. (I still have nightmares about yellow text on black screens).

    When I staffed the ER at the VAMC, and a patient rolled in, I could have just about everything I needed to know within a few taps on the keyboard, as long as I had his “last four” digits of his SSN. I didn’t need to see his real-time Doppler echocardiogram in living color, I just needed the REPORT. I didn’t need a 600 page document from his observation admit for non-cardiac chest pain last year, I just needed the result of his nuc stress test done 2 mo ago. (BTW, that 600 page document included all the student dietitian’s notes, given by the computer the same weight and space in the system as the cardiothoracic surgeon’s operative note for repairing his aortic stenosis!!!!)

    I could even quickly graph out the veteran’s creatinine over the last 3 years if I wanted, it took only a couple of clicks to select the lab parameter and tell the computer to graph it.

    Now, doctors are reduced to being only scribes. If I had known I was going to be just a typist when I went to medical school, I would have chosen being a secretary instead. What the VA (and other EMR vendors and proponents DON”T realize) is that doctors are literally spammed to death by EMR’s now.

    A few years later, I served as medical director for a large correctional facility. The prison inmates would often demand that we “get their records from the VA.” We would gladly acquiesce, but the records would literally have to be delivered in a moving van. One patient’s records would fill tens of thousands of pages of printed text, almost entirely garbage. Imagine a crate of standard typing paper, which holds 10 reams of 500 sheets each. There would be 3-4 CRATES for each patient that someone had to sift through, looking for the obscure operative note, discharge summary, etc.

    I had no desire to see the LPN’s computer-generated, template driven, 5 page report on the patient’s bowel movement on a hospital day 13 of his psych admit in 2004.

    What we need now is a system that resides “in the cloud” (though it makes me nervous) – and is SIMPLE, like the old VA monochrome system. If every Medicaid patient in the US had their data loaded into that system, and it was accessible to anyone with a legitimate DEA number, we could REALLY have the benefits of “EMR.” Why the government, in its infinite wisdom, can’t see that is beyond me.

  2. Yes, please read the report as the blog is in response to it’s findings based on an extensive set of interviews and research.

    Your experience with EPIC sounds great. The usability often is directly related to system build and user engagement in the content design. EPIC can work very well, but at what cost?

    Your comments on VistA are counter to this RAND report and to the many studies done on the system (and to our extensive direct use of the system). Recent physician surveys done by Medscape and the AAFP show it preferred to EPIC and all the enterprise EHR’s.

    But, your comments do reflect the commonly expressed “conventional wisdom’. But, the fact is that as far as quality outcomes, affordability, clinical adoption and interoperability are concerned. EPIC’s weaknesses are VistA’s strengths. The question they almost ask is: Is EPOC worth its excessive cost.

  3. I have to disagree completely. I was exposed to Epic during Kaiser Northern California’s rollout of HealthConnect in 2006 and, as a locums, got virtually no training. It did definitely slow down the process of seeing patients, during the roll-out. I think 20% is a reasonable estimate. On the other hand, templates could be tailored easily, readability was excellent, and the interface to the patient check-in system was terrific. Eprescribing was easy, with meds easily searchable on partial strings.

    If the ultimate slowdown is only 2 minutes, however, that’s something to rejoice about, because it means only an hour of additional work daily for 30 patients. I recently worked for a system whose CEO is making a national name for himself on the backs of his staff and patient population by inflicting VistA on them. His stated rationale is that it’s “free” and open-source. The system spews garbage and templates can only be changed (even when the IT department made a coding error) by aegis of the director of IT. It cannot data mine in rows, so every field has to be a new line. Notes are pages long because they’re in single column format. Free text breaks formatting if you insert anything after typing a paragraph. Meds are not searchable on partial strings, making CPOE a rousing game of “what am I thinking?” For example, trying to order vicodin brings up only “vicodin non-formulary.” Hydrocodone brings up nothing. You have to know to search for Norco, which brings up “APAP/ hydrocodone.”

    It’s so bad that almost anyone seeing high volume has resorted to using a scribe, which in this case is mostly non-college grads (premeds or high school grads hoping to go to nursing school) padding their resume and lacking medical terminology or writing skills such as grammar or basic spelling. The third time I had to correct “uterine thyroids” I gave up. Editing scribed notes is another waste of time. Understand that the system had already been using VistA for several years before I arrived, although not the CPOE component.

    I wonder how many of the physicians polled actually had experience with all 3? For the record, I worked briefly at another facility during its rollout of McKesson CPOE, which was a replacement for another EHR. All of the docs there had used the other system, as had the nurses. It was such a disaster that within a year, they became a captive of another larger system in order to piggyback on their EPIC license.

    My point is this: most VistA users have been in the VA system for years and it is familiar. They may not know any better. Most Epic and McKesson users are having to change from analog systems, which have greater user satisfaction for clinicians. I am in an unusual position, having used all three.

    I admittedly have not read the RAND study, but will do so now.