The Military Health System gave its providers an opportunity last week to sound off about AHLTA, an electronic health-record system originally deployed in ambulatory care but also intended for inpatient use.
The brass got an earful.
Long-held and pent-up user frustrations with the system—development work on AHLTA and its precursor, the Composite Health Care System II, began in 1999—came out in a gusher of 209 comments posted on a special Military Health System Web page set up for the global town hall meeting called by S. Ward Casscells, assistant secretary of defense for health affairs.
Since April 2007, Casscells, a physician, has served as the head of a healthcare system that includes 65 hospitals and 412 medical clinics that provide healthcare services for 8.6 million people, including active-duty service personnel, their dependents and retirees. Since its inception in 1999, defense, intelligence, aerospace and information technology contractor Northrop Grumman Corp. has been the primary systems integrator and developer of AHLTA.
An overwhelming majority of comments posted about AHLTA were negative, and some were vitriolic.
One user called AHLTA "an embarrassment," while another noted the system is "slow at its baseline operational status," but "when it goes into fail mode, which occurs on a weekly basis, it is intolerable."
"As an AHLTA user for over five years now, I remain completely disappointed," wrote another commenter. "AHLTA was designed for administrators—not clinicians—and it's slow, inefficient, unreliable and in every respect, an inferior product compared to other commercially available" EHRs.
Not that AHLTA had no supporters at comment session, with some saying many of the problems can be attributed to poor change management, a problem that has afflicted many IT system deployments in the private sector as well. But almost all of those who said nice things about Northrop Grumman's baby also found some devils in the details.
"Although AHLTA is frustrating at times," one person wrote, "it has been consistently shown that a majority of our problems occurring in the field are consistent with the customization and modification of local workstation configurations and the addition/use of software applications which absorbs resources."
Still, negative comment dominated the Web site. Several wrote out multiple paragraphs of complaints and one contributor posted the following numbered litany of AHLTA woes:
- "AHLTA does not give PCM ability to 'mine data' directly.
- AHLTA does not allow complete and uncompromised 'medication reconciliation' therefore remains a patient-safety risk.
- AHLTA Wellness modules do not work. They will not capture data for HEDIS (Healthcare Effectiveness Data and Information Set) measure compliance in many instances unless done at the MTF (medical treatment facility). This tool does not therefore assist in promoting population wellness.
- AHLTA does not communicate securely and seamlessly with other IT tools such as e-mail to patients.
- AHLTA impairs efficiency in providing patient care and should be abandoned. There are better products on the market now."
Another provider got specific about productivity losses attributable to the system.
"Prior to AHLTA, I was able to provide over 5,000 visits a year with excellent notes. Now I am lucky to reach 3,600 per year with a note that is cryptic in nature."
Multiple posters compared AHLTA—almost always unfavorably—with the Veterans Health Information and Systems and Technology Architecture, or VistA clinical IT system at the Veterans Affairs Department.
One wrote: "Please explain why the DoD rejected an award-winning, highly praised health IT system known as VISTA in favor of AHLTA, which to date seems dysfunctional?" Another agreed with that poster, adding, "What on earth was wrong with VistA, the VA system? It is an excellent system, much more user-friendly. This system, since it is not available often, leads to me missing pertinent information in the medical record, and potentially bad outcomes." Yet another wrote: "I would strongly suggest that the DoD consider switching systems to the VA system."
Others called for AHLTA to be replaced without specifying its successor. According to one, "I strongly believe that AHLTA is grossly inadequate and is not able to be repaired. My only real question about AHLTA is: When will the DoD scrap it and find a functioning system?"
Switching to a different system is not likely to happen, however, according to Terry Jones, a Defense Department spokesman.
"Dr. Casscells has made this point very clearly as nearly as yesterday (Monday); we're not scrapping AHLTA," Jones said. "That's not part of the equation. What they're talking about doing is taking AHLTA and VistA and making the two systems better."
According to Jones, that is not to imply the Military Health System is whitewashing the problems or that Casscells is ignoring the complaints.
"He was not surprised by the tenor of comments," Jones said. "Dr. Casscells has said that AHLTA is difficult to learn, and once you do, it's cumbersome and difficult to navigate. A big challenge to making it more user-friendly is making it more interoperable with the VA's system, VistA. We're expecting sometime later this week or perhaps next week to come back with responses."
Last August, national intelligence, IT and management consultant Booz Allen Hamilton contracted with the Defense Department and the VA to study the feasibility of creating a common EHR system for use by both of the two service-oriented healthcare systems.
"They got the report from Booz Allen on Friday," Jones said, adding: "I don't know what they are recommending because I haven't seen the report. A big part of this is what to do with AHLTA and VistA. But there is a need to upgrade both systems, and there is a need to work out what Dr. Casscells terms 'a convergent evolution of AHLTA and VistA,' " Jones said.
Efforts to make the Defense Department and VA systems interoperable date back at least a decade to a 1997 executive order by President Clinton. Since then, however, the Government Accountability Office and its predecessor, the General Accounting Office, have issued reports or had staffers testify before Congress at least 10 times about the lack of progress toward that goal.
The schism between the Defense Department and VA clinical IT systems actually can be traced back nearly a decade earlier to 1988. That's when the Defense Department issued a contract to create its own clinical computer system, paying more than $1 billion to outside contractors headed by Science Applications International Corp. to convert the VA's free, public domain Decentralized Hospital Computer Program (later renamed VistA) to create CHCS I, a proprietary system for military use. But that effort begat inoperability between the Defense Department child and VA parent systems, a problem that has not yet been completely remedied even at the expense of hundreds of millions of dollars.
Northrop Grumman is also the prime contractor for two of those interoperability efforts, the joint Clinical Data Repository/Health Data Repository and the Bidirectional Health Information Exchange. In March 2007, it was awarded a contract extension for systems integration and support for AHLTA. The nine-month contract was for $67.7 million with two one-year extension options, the company reported. It also was selected by HHS as one of four prime contractors to develop a prototype of the proposed, civilian national health information network.
Joseph Dal Molin is vice president of business development for WorldVistA, a not-for-profit organization formed to create an open-source version of VistA for use by healthcare organizations outside the VA, both in the U.S. and abroad. Dal Molin said that he is working with the government of Jordan to adapt the VistA system for use in the public health system of that Middle Eastern nation.
Dal Molin said that AHLTA "is sort of a byproduct of where the DoD is going in its culture with outsourcing and purchasing."
According to DalMolin, VistA, despite its success within the VA, gets a bad rap by outsiders because it is based on the Massachusetts General Hospital Utility Multi-Programming System, or MUMPS, a 1960s-era database and programming language developed initially for healthcare, but now disparaged by some in the IT community to be "old technology."
At the Defense Department, Dal Molin said, "The technical decisions aren't being made from a clinical point of view, but whether this is a better architecture, and not at the point of the practice of medicine. The technology is driving the decisions instead of the evidence and the outcomes. It would be a major effort to switch over to VistA, but ruling it out for technical reasons is not looking at the evidence and weighing the evidence."
The software development process the VA uses to create the VistA system, a distributed development effort by which modules of the system were created in a close, iterative collaboration between developers and clinicians, also is underappreciated, he said.
"VistA needs to be improved—there is no argument about that—but the processes the VA used to get to VistA, internal collaboration and open development, have seemed to be much more effective when you get to a scale of the VA or DoD, than outsourcing," Dal Molin said.
Several commentators on the Defense Department Web site agree that a collaborative process is best, including the one who said, "It seems that AHLTA was designed from the database engineer up rather than from the practitioner down. That forces the doc to practice the way the engineer thinks rather than the obverse. That's absolutely backwards."
What do you think? Write us with your comments at [email protected]. Please include your name, title and hometown.
