GovernmentExecutive.com

Return to Article: A New VistA for AHLTA?

• 84550
  

  I have been on the AHLTA project for seven years and have watched providers and staff use the system in spite of the problems. I have come to the conclusion that in spite of the problems, AHLTA still affords the providers and the patient with an Electroinic Health Record which enhances patient care and make the patient's medical conditions readily available. AHLTA represents a new era of medical documtnation.

  james r. Worthy Posted August 11, 2009 2:49 PM
• 59349
  

  AHLTA difficult to use? Hard to train users? You havent met an Air Force AHLTA Transition Support Trainer. My colleagues train users to take advantage of what does work in AHLTA, but the mere fact that a user can document on day one and get that medical information into the database is certainly remarkable from any standpoint. Anyone with a global perspective knows that AHLTA and any program like it would face the same challenges integrating and implementing into the current standard desktop configuration. The idea is sound, the idea is viable and sustainable. Our infrastructure should evolve as well, and we certainly should regress to another documentation tool that may or may not perform as well as AHLTA. I agree with users who bear the brunt of all the changes or lack of change in the software developement. I am looking forward to an AHLTA that delivers wihtout all the superlatives from the company that owns it. I want AHLTA to work more efficiently as well and so do my colleagues.

  JEFFREY FALLANG Posted September 25, 2008 9:08 AM
• 55123
  

  There's another approach to the transition from "the antiquated technology" referred to by Dr. Casscells to the prevailing technologies of Java and Relational Databases that so far has not entered the debate on future direction that is presently taking place within the VA and MHS. An Israeli software company, CAV Systems Ltd, has developed technology which automatically converts MUMPS systems into Java/Oracle systems (or any other Relational DataBase Management System including open-source systems such as MySQL and PostgreSQL). The converted system is functionally identical to the original system. The look-touch-feel is the same. Furthermore, the converted system can continue to be maintained and enhanced using either MUMPS or Java without the need for any MUMPS software in either the development or deployment environments - amazing but true. This is "game-changing" technology that opens up a wide variety of roadmaps to the future of AHLTA/CHCS/VistA that no other approach can offer. This is not vaporware - and the evidence is already known to key players within the relevant government agencies. Might be worthwhile for the powers-that-be to take a closer look at this technology and its ramifications.

  Disclosure: the author of this comment is VP Corporate Development of CAV Systems Ltd.

  Alex Hill Posted July 28, 2008 2:18 PM
• 54920
  

  I've worked in various technical and training capacities out at the Military installations in the implementation of AHLTA. My peers on the project(s) would agree that it's been a rough 6 years. We've stood before angry and frustrated Doctors, Nursing and Support staff and defended a system that would've had a reasonable chance to work anywhere else, but in the context of Service Branch differences, contract turnovers, perpetual BETA Code conditions, questionable testing, poor underlying network infrastructure and verying degrees of Command support, it isn't surprising that the users were cynical.

  Then, throw in poor performance. Software folks: imagine glomming a centralized DB-GUI-Client Server platform on top of a decentralized DB, text-based system, constantly updating each other, thrashing tens of millions of records. Network folks: imagine maintaining multiple host servers, worldwide, communicating to multiple cache servers and a central repository with massive layers of milspec security.

  With fifteen minute routine visits, you need to help the Docs make a 3 1/2 minute note. I don't remember hearing SLAs mentioned at all.

  Leon Stafford II Posted July 24, 2008 10:51 AM
• 54834
  

  Back when MHS decided that they had to migrate CHCS I, they funded a couple projects to show proof-of-concept for the different approaches. The CPR Interoperability using Object Oriented Technology (CIOOT) group was formed from Battelle Memorial Institute and ESI Technology Corporation led by Dr. Janet Martino. In a nutshell, the goal was to save the data and data structures on the existing CHCS I system (MUMPS systems) and replace the rest of it with a modern OO system that could be evolved overtime. ESI had evolved the MUMPS technology to a modern OO environment (EsiObjects). We built a set of tools that migrated the existing File Manager "files" to Object Oriented interfaces (objects). We did the conversion on the entire CHCS I system at that time. Battele built a modern GUI for demonstration. At the time they used CORBA as the middleware to ultimately integrate the CHCS I systems to solve the problem of distributed patient records. This approach would have totally modernized the CHCS I system by evolving it, saving the data (no conversions) and minimizing the cost. Obviously, it was rejected and CHCS II was given the contract.

  Later, the CHCS I object data wrappers were used in a project at the Naval Medical Center San Diego (NMCSD) where, under the direction of Dr. Emory Fry (CMD), they were provided access to CHCS I data for a Web application. Additionally, in another project, ESI used the tools to generate object interfaces to all VistA "files". We then showed that patient records from CHCS I and VistA could be presented as an integrated record using the Web Browser. It was presented and rejected.

  Recently, a commercial project was completed to implement a new Registration module for the World VistA version. A new Web Services interface to VistA was built using the EsiObjects technology. The front-end and middleware were built using existing, modern technology (Java, etc.)

  Bottom line is that technology is not the problem - it was there and proven 10 years ago. The concept of the "antiquated technology" (MUMPS systems) has evolved to modern technology and is not longer an issue.

  Oh, by the way, all of this is free - it's Open Source.

  Terry L. Wiechmann Posted July 23, 2008 2:29 PM
• 54778
  

  As one of the early software architects of both VistA (DHCP) and CHCS, I have a 30 year history in dealing with VA/DoD health care integration. We had a meeting in Oklahoma City in December, 1978, with VA, DoD, and IHS folks. I had always assumed that the same toolset (FileMan, Kernel) would be applicable to all agencies. While working for the VA, I had a working interface between March AFB and Loma Linda VAMC in 1984. While working for SAIC, I did two more interfaces in the 1990's. The problem is not with the technology, it is in the organizational turf wars.

  Perhaps it is time to consider a fresh alternative - a fully patient centric architecture in which patient information exists in a "cloud" that is not specific to any particular agency or enterprise.

  The DHCP architecture was based on an amazingly simple approach (one language with 19 commands and 22 functions) providing the simple initial conditions that allowed the system to evolve gracefully in concert with users' needs. It was designed to evolve and adapt, not be driven by top-down specifications. Rather than call VistA antiquated, I think it merits intense study and reflection as to why it has endured over the years.

  Tom Munnecke Posted July 23, 2008 12:51 AM
• 54749
  

  The effort and risk involved in replacing a system based on a signficant investment by the US Government over many years is not to be taking lightly. I think it is important to be very careful when looking at the available information about both the Department of Defense systems and the Department of Veterans Affairs systems as both are multi-faceted with significant history behind each of them. CHCS I is currently still used in many military care institutions, and should not be confused with AHLTA. I personally have no experience with AHLTA, but have worked with both VistA in the present, and CHCS I in the past. Beyond CHCS I and AHLTA, there are many specialized medical applications within each branch of the military, many of which connect to CHCS I. Both VistA and CHCS I are based on the Decentralized Hospital Computer Program (DHCP) of the 1980's. The quoted statement by Mr. Casscells who said VistA, developed in 1994, is based on "antiquated technology" seems to be confusing the age of technology with the reliability and efficacy of the technology in supporting the delivery of quality health-care. Regardless of the age of the technology, the use of VistA has been documented numerous times as positively reducing the incidence of medical problems and errors. The literature is very clear that the use of VistA has been a wise use of government funds and has helped veterans many times over. The concerns of the MHA professionals are significant, but don't please confuse ease-of-use issues with quality of record keeping issues. I'm confused by one previous comment, as I know from experience that there are many fine professionals in the VA who understand the issues of modern software development and work daily to support health IT. I applaud any and all efforts to use modern computer methods to enhance safe and affordable health care.

  David Whitten Posted July 22, 2008 1:06 PM
• 54701
  

  Sorry if i misstated CPHS Vista's outpatient capability. The VA has reliable, easy -to-use electronic record systems.But as Drs Peake, Tibbets, Kussman and Chuck Hume have stated, VistA technologies need to be upgraded. DoD's AHLTA is hard to learn and use, slow, and often down. Most experts tell us that the risks to patients, and dollar costs, of replacing AHLTA with VISTA would be too high, that there is not yet a plug-and-play commercial system that we could buy. So we may well end up trying to upgrade both VistA and AHLTA in coordination, aiming at convergence but at least assuring increasingly detailed and reliable 2-way exchange of data, to the point of real interoperability. The final product is not likely to be very similar to today's VistA or AHLTA, but most clincians like the look and ease of the VistA GUI . We will also be offering a choice of personal health records. We have some serious congressional deadines to meet, and Chuck Campbell's team, which reports to Principal Dep ASD(HA) Steve jones,PhD, is scrambling to sort out consultant input, doctor input, and bring me a proposal this week. I will review that with our Surgeons General, Dr David Chu, DepSec Gordon England, and with SECVA Jim Peake and VA's top doc UnderSec Mike Kussman, and VA IT leaders Tibbetts and Hume. We also coordinate with HHS SEC Leavitt's AHIC program led by ONC's Bob Kolodner, MD, to help shape - and comply with - national standards. We are grateful for the advice, and for the dedication of the docs who have struggled conscientiously with a frustrating system. I am sure we will make it better, though the pace will not be as fast as any of us would like given the size and complexity of these agencies, the legacy systems, pre-existing contracts, and the checks and balances of govt. Keep up the fire. Ward Casscells,MD

  ward casscells Posted July 21, 2008 10:43 PM
• 54674
  

  I believe Dr. Casscells has been misinformed based on his statement that VistA has no outpatient system. VistA does definitely collect outpatient healthcare in the Patient Care Encounter (PCE) and it is consolidated in the enterprise wide National Patient Care Database (NPCD).

  Mike Leigh Posted July 21, 2008 1:37 PM
• 54660
  

  Having worked in the MHS Medical IT field since 1992 (prior to CHCS I), I agree with the clinicians. However, I think a design team internal to the VA and MHS who clearly understands the business/clinical processess and the IT (good database design architecture, UI separation from the data processes and DIACAP security) needs to be formed to address this issue. Using the current form of VistA or AHLTA will not work. Trying to morph the two together will not work. Designing and building a system with data connectors to both legacy systems will work using an intelligent distributed data structure with SEPARATE OLAPing and OLTPing parallel structures.

  Stephen Wolfe Posted July 21, 2008 10:54 AM