Page 1       TOP OF DOC

House of Representatives,
Subcommittee on Oversight and Investigations,
Committee on Veterans' Affairs,
Washington, DC.
    The subcommittee met, pursuant to call, at 9:30 a.m., in room 334, Cannon House Office Building, Hon. Terry Everett (chairman of the subcommittee) presiding.
    Present: Representatives Everett, Clyburn, Snyder, Mascara and Evans.
    Mr. EVERETT. Good morning. We will come to order.
    Today's hearing will examine if the VA computer systems will work or fail after 12:00 a.m. on the morning of the year 2000. Many computer systems we use today use a two-digit date to recognize the year. With a two-digit format, computers will fail to operate correctly, because the year 2000 will read 1900.
    GA warns that the payments to veterans with service connected disabilities could be severely delayed, because VA's compensation and pension systems either halt or produces checks so erroneous that the system must be shut down, and the checks must be processed manually.
    Time is certainly running out. Our computer generated display on your left is a real time countdown which comes from the Internet. It shows how much time is left before January 1, 2000. Our low technology back-up display which we prepared in case the computer system went down shows how many calendar days, 918, are left.
    Will VA's computers crash or shut down, malfunction or compute incorrect information? How widespread of a problem will this be? What will be the cost of these errors? Will health care be delivered to our veterans safely and without interruption? Will their paychecks and pension checks get their safely without interruption?
 Page 2       PREV PAGE       TOP OF DOC
    We'll examine VA's efforts since they testified on their modernization efforts almost a year ago. It is interesting that the Veterans Benefits Administration made the year 2000 Y2K problem a priority only last June.
    Today we will hear from our colleague, Steve Horn, Chairman of the Government Reform and Oversight Subcommittee on government management information and technology. We will also hear from the GAO, which has done extensive analysis on the Y2K problem, governmentwide and in the VA.
    The VA will explain what they have done to address Y2K. Tom Shope of the Food and Drug Administration will tell us what actions the FDA is taking to address Y2K issues, to ensure the safety and effectiveness of medical devices in the health care industry, including the VA.
    I think we have a pretty full plate for discussion today, and I look forward to hearing testimony from all our witnesses. I will now recognize the ranking member, Mr. Clyburn.
    Mr. EVERETT. I will now recognize the ranking member, Mr. Clyburn.
    Mr. CLYBURN. Thank you, Mr. Chairman. I am pleased to join with you today in calling for this extremely critical hearing on the VA's efforts to achieve Year 2000 compliance.
    The GAO tells us that the VA has a long way to go to solve this problem, but not much time to get there. I am encouraged that this subcommittee has decided to place a watchful eye on the VA's progress in this regard. I am hopeful that, through continued oversight by this subcommittee, we can help to ensure that the VA is able to achieve Year 2000 compliance.
 Page 3       PREV PAGE       TOP OF DOC
    The purpose of this hearing is to hear a status report from the VA on their progress on this critical issue, to underscore our subcommittee's interest and concern, and to make clear this committee's expectation that our deserving veterans will receive uninterrupted benefits and services in the year 2000 and beyond.
    I believe it is important to recognize that this is not a problem that is unique to the VA. As a recent cover story in Newsweek magazine points out, this is an issue that effects everything from the personal computers many of us have in our homes to private business and industry and to nearly every local, State and Federal government across the globe.
    Mr. Chairman, I would ask unanimous consent that the June 2, 1997 Newsweek cover story titled ''The Day the World Shuts Down'' be included in this year's hearing record.
    Mr. EVERETT. Without objection.
    [The attachment appears on p. 37.]

    Mr. CLYBURN. It is also important to recognize that the VA has been working hard over the past several months to get their act together on this issue. I want to commend these recent efforts and want to make it clear that I do not doubt the sincerity of the VA's interest in addressing and ultimately solving this vexing problem.
    I must say, however, that the objective views of knowledgeable outsiders strongly suggest that the VA's task is more daunting and difficult than its written testimony to this subcommittee might seem to suggest.
    I would also like to note for the record this morning that the Gartner Group, a leading independent industry group, has conducted extensive research into the Year 2000 problem and has monitored the steps private industry and government have taken to address the problem.
 Page 4       PREV PAGE       TOP OF DOC
    Unfortunately, scheduling difficulties prevented the Gartner Group from providing live testimony before our subcommittee today. The Gartner Group has graciously volunteered, however, to provide written responses to any questions any of the members of this subcommittee may have.
    Mr. Chairman, I ask unanimous consent that subcommittee members be allowed 5 business days to provide the committee with written questions to the Gartner group and that the responses be included in the formal hearing record.
    Mr. EVERETT. Without objection, so ordered.
    Mr. CLYBURN. Thank you. Mr. Chairman, as I stated at the outset of my remarks, the GAO tells us the VA has a long way to go to achieve Year 2000 compliance, and the clock tells us they don't have much time to get there.
    Through continued close scrutiny and oversight by this subcommittee, I am hopeful that we will remain committed to doing what we can to help ensure the VA makes it to the year 2000.
    Thank you, Terry, for your leadership on this issue.
    Mr. EVERETT. Thank you, Jim.
    [The prepared statement of Congressman Clyburn appears on p. 33.]

    Mr. EVERETT. Mr. Mascara, do you have any comment?
    Mr. MASCARA. Yes. Thank you, Mr. Chairman, and thank you for calling this hearing this morning to examine a serious VA computer problems.
    Last evening I read over the material provided by the committee, and I must say I am most alarmed that, if this problem is not corrected, and corrected quickly, it could result in late benefit checks and denied benefits to millions of veterans across America, a situation which this subcommittee cannot tolerate.
 Page 5       PREV PAGE       TOP OF DOC
    I understand that, since my colleagues on the House Committee on Government Reform and Oversight gave the VA a D for its efforts to correct this problem and this hearing was scheduled, the VA has begun to move. Those in the know, however, say that the plan being developed by the VA is very general and raises more questions than it answers.
    I am hoping that today's testimony will help alleviate the committee's concerns and not give us cause for further heartburn.
    The bottom line is that too much is at stake here to even think of the VA not getting the problem corrected. Veterans are relying on us to keep the pressure on the VA, and you can all rest assured we will do exactly that until this problem no longer exists.
    Thank you, Mr. Chairman, and I yield back the balance of my time.
    Mr. EVERETT. Thank you very much.
    [The prepared statement of Congressman Mascara appears on p. 45.]

    Mr. CLYBURN. Mr. Chairman, before we get started, I think it may be worth noting that this is Mr. Mascara's first day as a member of this subcommittee. I want to welcome him to the subcommittee.
    Mr. MASCARA. Thank you, Mr. Clyburn.
    Mr. EVERETT. We certainly do. We do welcome you, and we need all the help we can get.
    Mr. MASCARA. And I'm delighted to be a part of it.
    Mr. EVERETT. I've got great help right here, but I always welcome more.
    Now, of course, I'd like to welcome my colleague, Steve Horn, not only a colleague but a classmate. He's Chairman of the Government Reform and Oversight Subcommittee on Government Management, Information and Technology.
 Page 6       PREV PAGE       TOP OF DOC
    Steve, in the beginning I would say that this is an elusive target. I started first having hearings on this as Chairman of the Compensation, Pension, Insurance, and Memorial Affairs Subcommittee 2 years ago. I think perhaps there's been some movement. We, frankly, though, are not near where we need to be, and I'm certainly going to be interested in hearing your testimony this morning, if you will please begin.
    Mr. HORN. I thank you very much, Mr. Chairman. I'm delighted you and I have been colleagues, and they picked the right person for the subcommittee chairman, since a publisher who knows what an investigation is is certainly worthwhile presiding over one of these groups.
    I'm going to skip-read through some of my testimony. I assume that's put in the record at this point.
    Mr. EVERETT. The complete testimony will be entered into the record.
    Mr. HORN. Let me just say a few general things, since you're into this subject. The year 2000 problem, when our Subcommittee on Government Management got into it, the Federal Government had hardly paid any attention to it in terms of the Executive branch. A few State governors were working on it, Governor Ridge in Pennsylvania, a former colleague here, Governor Wilson in California, and some others.
    They brought in chief technology officers to pull things together. So the States were a little ahead of us, and we thought we would see what's going on in 24 departments and executive agencies, and I believe you have the results before you.
    Counsel for this study on my right, Mark Uncipher, has been with this for a year and a half now, and our 16th report of the full committee is about this subject. So I commend it to you and the staff, and our 24 agencies and the grades we gave them are in there.
 Page 7       PREV PAGE       TOP OF DOC
    The Veterans Affairs Agency got a D. There were four basic questions we asked at that time: Has the agency a Year 2000 plan? The answer out of Veterans Affairs was no. Is there a Year 2000 program manager that's been appointed to get some direction and focus in this area? Now Veterans Affairs did say they had that.
    Now does the agency have any cost estimates for the Year 2000 solution? No, big blank. Did you answer the rest of the questions we asked, which were about 11 or 12 other questions? They did answer those.
    When we went through this, we only had four A's in the whole Executive branch, Agency for International Development, Office of Personnel Management, Small Business, and Social Security.
    There were three B's, three C's, 10 D's, four F's. Being a former professor, I grade on the absolute. I do not grade on the curve. So the D's and F's in many universities would have become C's and B-'s, but not here.
    Now that shook them up a little. We had two cabinet secretaries that never even heard of the problem. One was the Secretary of Energy, and I figured she wouldn't have heard of the problem since she was traveling so much, and we were doing a separate investigation on that.
    The one that surprised me was the Secretary of Transportation at that time, and he's a very able person, and he hadn't heard about, and we didn't know at that time what we know now, that one of his key agencies in the Transportation, the Federal Highway Administration, had been working on this problem since 1989. They were a first, along with Social Security, that's been working on the problem since 1989.
    We look to Social Security to be sort of the guards, tackles, everybody else that's running down the field ahead of the other agencies and, hopefully, a model; because, as you and I know, if Social Security can't get this problem solved by January 1, 2000, there will be 435 district offices of members of the House that will have a few hundred thousand people saying where's my check, where's my eligibility. This is a very serious question.
 Page 8       PREV PAGE       TOP OF DOC
    When we started people thought, oh, well, you know, what's this all about, science will solve it. Science hasn't solved it. Now all of this happened, as we know—and I remember using those computers in the university 20-25 years ago. We had very little storage capacity, and somebody had the bright idea, hey, we can save a few things here and there. Instead of having 1966, let's just put in 1966.
    So you put in 1966, and you can't get the whole thing in. So when you get to the year 2000, you got two zeroes there, and the computer doesn't know what to do. Sounds crazy, but it's there.
    Nobody has solved the problem yet. If there is a simple solution, they will be a billionaire overnight, because as Gartner testified when they appeared before us, a consultant group, fairly widely recognized, it's a $600 billion worldwide problem.
    We're half the computers in the world. It's a $300 billion U.S. problem, private, nonprofit, governmental. The Federal Government, they estimated, was a $30 billion problem. My instinct said I don't think it's that much, but it's something serious.
    As I ought to say, I'm the least knowledgeable person probably in the Congress on computers. So don't take my instincts as gospel, but the fact is that we asked through the Appropriations Subcommittee for the budget director, in submitting the President's budget for Fiscal Year 1998, to give us an estimate of what they think it is.
    Now Mr. Raines has done an excellent job in getting the Executive branch to take this issue seriously. He's the first one that really ran with it, and he also, I think, agrees with what I've been saying. Don't send us a big budget request and waste a year up here going through the process; reprogram money, and get on it now.
    He has taken exactly that philosophy, and the people in the Executive branch have received the message, get with it, don't just whine about it. I think that is very important.
 Page 9       PREV PAGE       TOP OF DOC
    Anyhow, he got into it when he was nominated and confirmed for Director of Office of Management and Budget, and he started getting some quarterly reports from them. Then he put up his estimate for the budget for Fiscal Year 1998, and generally it was $2.3 billion. Well, I started laughing at that one. That was just plain too low, and my instincts, I think, are right on that; because I then held a hearing.
    Assistant Secretary Paige, General Paige of Defense, said, well, $1 billion of that 2.3 of his is from the Department of Defense, and we haven't even started assessing anything. Now the Department of Defense has some very complex problems. They don't even know the extent of them yet, but they are working on it, and I suspect when this gets all sorted out, it will be somewhere between $10-15 billion problem.
    Now there is a timetable that the Director of OMB has set and, as you look at it, you start worrying, because some of it is pushed right up into 1999. Now anyone that's ever dealt with massive computer purchases and implementation—and I have, both as a university president and as an oversight chairman where we've had such wonders to look at as the IRS's $4 billion mess, and when I was on the Aviation Committee chaired by Mr. Oberstar in the 103rd, he took a couple of us out to look at the FAA boondoggle, and you could walk in the room and see that it was a boondoggle, and they were at the $4 billion point, too.
    I've asked both those agencies, why can't you learn something at the $4 million, $40 million, $400 million point? Why does it seem to have to go to $4 billion. It's because they have had poor management, and everybody that had a bright idea overnight starts changing the thing, and there was no focus.
    You can't be up to the latest point in time. You're always going to be behind in the advance of technology, and they need to realize that.
    So what we have been after is working with OMB in a sort of joint questions that go out. They share them with us. We just have the new batch in. We will be grading those and probably releasing it sometime in July. We want to take a very careful look and be very fair about it, because I think a lot of the agencies do now have the message.
 Page 10       PREV PAGE       TOP OF DOC
    I think, when you look at the Veterans one, as I mentioned, they really only had two of four base categories the last go-around, and we would hope by that time that there would be progress; but I also find the comments contained in the GAO testimony that you're going to hear today very troubling.
    I don't want to steal the General Accounting Office's thunder, but I'd be concerned that there is not yet a complete inventory of local computer applications, that there's not a better system for prioritizing mission critical applications and that there are not contingency plans for systems failing, and they inevitably will.
    I've seen this across America in universities that had millions spent on student registration. The whole system breaks down, and the poor little old ladies that have been running registration for years get called back from retirement to go back and do it the old way where you stand in lines for 3 days.
    So we need contingency plans that work, and I think that's about it for the once over lightly, and I commend you for keeping after the agency under your jurisdiction.
    [The prepared statement of Congressman Horn appears on p. 52.]

    Mr. EVERETT. Thank you very much, Steve.
    Before I go into some questions, I'd like to mention we've been joined by the full committee ranking member, Lane Evans, who was also the very able chairman of this committee before it was disbanded a couple of years ago. His knowledge and expertise and the desire he has for our veterans is well known here on the Hill.
    Steve, you bring up several interesting points, and one that I discussed a little bit with VA is trying to nail down source code, where it is, who has it, who has the ability to write it. From what I've been able to gather, source codes are changed in a number of different regions and places.
 Page 11       PREV PAGE       TOP OF DOC
    That has led partially to the problem that we have now in trying to, number one, find where the source code is; and, number two, once you get a number of different people changing source code on a program, then you have people working at cross-purposes.
    Seems to me, it would be much better off—I don't know if this exists in other agencies or not, but it would be much better off when we approach solving this problem if we centralize where these changes can be made rather than having, as you mentioned a minute ago, anybody that came up with a bright idea made a change.
    Would you have any comment on that?
    Mr. HORN. Well, not really. I think you need to get in a panel of experts, but I do know this, that when you have millions of lines of code that you've got to work your way through and ask the question, does this agency need this to carry out its operations—and Social Security would be a good group to have in. They've worked with this.
    They have millions and millions of lines of code. They estimate at Gartner why they got to $600 billion was that it would be about the cost of a dollar to look at each line, when you take in the wages of the people involved.
    In solving this, this is not simply a money problem. This is a human resource problem, because as we get nearer and nearer to 1999, there's going to be a lot of panic set in in the private sector, which we're also trying to alert, and nonprofits, local State governments, so forth, that might have been lagging along.
    That means the cost of those experts that know how to bring up the code, deal with it, and adjust it will be going up and up. That's why we might well have a $300 billion national problem here, but if we do this now in an orderly manner, maybe the problem can be solved without the scarcity of resources and talented people not being around to solve it.
    Mr. EVERETT. We are also concerned about the problem of embedded chips, particularly those with medical devices that may malfunction because of the Year 2000 problem. What more should Congress be doing to improve the awareness and response to this problem that, while it may be small, it could be critical?
 Page 12       PREV PAGE       TOP OF DOC
    Mr. HORN. Yes. I think you ought to get in a panel of people from the industry. We worry about the medical device problem and the embedded chip also, and that's where the assessment of these agencies is so very crucial.
    If you're dealing with huge hospital systems, as the services have and the VA has, that might compound your problem many times over, and we're just not aware of it.
    So all I can commend to you is go out and get some good people that know something about the technology to talk to the experts as they are grappling with these questions. As I say, time is the significant limiting factor here. The clock keeps ticking, whether we're acting or not.
    Mr. EVERETT. Mr. Clyburn.
    Mr. CLYBURN. Mr. Chairman, I have a question, but I wondered if I ought to yield to the ranking—I do have a question, Mr. Horn. Have you taken a recent look at the VA and seen where they are today as opposed to where they were when they received the D grade?
    Mr. HORN. Well, we have not come out with the latest. We did have the chief technology officer, chief information officer for the VA, before us in a panel; but I couldn't give you a grade at this point. We just haven't sat down and looked at the pieces, and I don't like to shoot from the hip.
    Mr. CLYBURN. I appreciate that, but I just wondering if you had taken a look at it.
    Mr. HORN. No. We are— —
    Mr. CLYBURN. I think I've made a B or two in my lifetime but, fortunately, I didn't stay down there. I was just wondering whether or not any movement had been made, according to your assessment.
    Mr. HORN. Well, I think all agencies are now moving after our little April 1996 hearing and the grading during the summer, but the question is how fast are they moving. We don't really have that information until we go through this latest report, but the key is pinning responsibility, getting organized, and starting to make the basic assessment.
 Page 13       PREV PAGE       TOP OF DOC
    You asked about the embedded chips. That's exactly the problem there. I understand that, really, the percentage of devices in equipment may be only 2 to 5 percent, and they may be affected. They may malfunction. However, in some pieces of equipment such as a medical device, with that being in there that's a catastrophic situation, and you might just not—you might overlook that, if you don't think about it hard.
    Mr. CLYBURN. Well, I want to commend you for looking into this matter, and I think that, from what I hear, you and ranking member Maloney have been very diligent, and I commend you for it.
    Mr. HORN. We've had 100 percent cooperation from everybody.
    Mr. CLYBURN. Thank you. That's all I have, Mr. Chairman.
    Mr. EVERETT. Thank you, Jim. I would say for the record that our colleague, Congresswoman Maloney, was scheduled to appear here and was unable to do so at the last minute, and we will welcome here testimony for the record.
    Lane? Oh, I'm sorry. We do come—first come, first served. All right, Lane.
    Mr. HORN. I want to say to the gentleman from Pennsylvania, we miss you, Frank. Why did you leave us? We could use you.
    Mr. EVERETT. Don't we need him?
    Mr. EVANS. Mr. Chairman, I want to thank you and the ranking member for holding this hearing and to thank our colleague from California for joining with us. We hope we'll continue to tag team with you as we progress, hopefully progress, on this issue.
    What's our next step? When do the next report cards come out?
    Mr. HORN. July. We want a chance to look at the data—the data has just come in—and really go over it thoroughly, perhaps send Mr. Uncipher and some others on the team out to some of the agencies.
 Page 14       PREV PAGE       TOP OF DOC
    Mr. EVANS. Thank you, Mr. Chairman.
    Mr. EVERETT. Thank you, Lane.
    Steve, we certainly appreciate your enlightening testimony. We also appreciate your dedication to this subject matter. It's not a very romantic thing that gets a lot of attention, but it's a most critical thing that could cause this nation great harm and great, great expense if we don't get the problem solved, and I share your pessimism in the way we're moving right now.
    Thank you for coming.
    Mr. HORN. Well, if every authorizing committee does what Veterans Affairs is doing, we don't have to worry.
    Mr. EVERETT. Thank you very much.
    Mr. EVERETT. I'd like to introduce our second panel, Mr. Joel Willemssen, Director, Information Resources Management, Accounting and Information Management Division of the GAO, and ask him to please introduce his panel.
    Mr. WILLEMSSEN. Thank you, Mr. Chairman, ranking member Clyburn, ranking member Evans, Congressman Mascara. Thank you very much for inviting us here today to testify on VBA's efforts to address the Year 2000 computing issue.
    Accompanying me are Helen Lew, Assistant Director, and L.J. Latham, Technical Assistant Director.
 Page 15       PREV PAGE       TOP OF DOC
    As requested, I will briefly summarize our statement.
    As with other agencies, VBA could face widespread computer system failures as the Year 2000 nears, due to the use of two digits to represent the year. More than most other agencies, however, VBA's failure would be felt by millions of people if the benefits and services on which they rely were disrupted.
    Eligibility for many of these benefits and services is date-dependent, which places their delivery at risk. Especially susceptible to disruption could be compensation and pension systems that relate dates to benefits, such as dates of birth or military service.
    VBA recognizes that the Year 2000 computing issue poses a serious challenge to them. Its information resources management plan clearly states that achieving Year 2000 compliance is the agency's number one priority.
    VBA has also initiated actions to assess its vulnerability and perform the modifications that must be made to its information systems. However, several substantial risks remain.
    In a report issued to you, Mr. Chairman, and being released publicly today, we detail these risks, and I'd like to briefly highlight a few of those.
    First, the structure of VBA's Year 2000 program management office needs strengthening, and technical and managerial issues need to be addressed. An agency level program office is needed to coordinate and manage the full range of interdependent information systems activities.
    A critical technical deficiency is VBA's lack of an overall systems architecture or blueprint to guide and constrain the development of replacement systems and the evolution of related systems.
    Second, much work remains to be done in determining whether VBA's information systems and their components are Year 2000 compliant now. VBA expected to have completed all of its inventories of systems by September 30, 1996. However, by that date inventories had only been completed for software applications at its three systems development centers.
 Page 16       PREV PAGE       TOP OF DOC
    According to VBA, part of the reason for the delay in completing inventories is the agency's loss of well qualified employees to retirement during recent agency buyouts. VBA's inventory also does not include local applications developed by regional offices.
    According to VA's own Year 2000 readiness review, without a complete inventory of regional applications, VBA cannot adequately predict or plan for the impact of the Year 2000.
    Third, VBA has not developed contingency plans for all of its critical systems. Three of its major business areas currently lack contingency plans to ensure continued operations in the event of Year 2000 failures.
    Fourth, VBA does not yet have sufficient information about the costs and risks associated with its Year 2000 activities. As a consequence, it lacks the information necessary to make decisions about prioritizing its information technology projects.
    Given the serious risks associated with VBA's Year 2000 activities, our report recommends that the Secretary take ten specific actions to help ensure the agency's success in making its systems Year 2000 compliant.
    In commenting on a draft of our report, the Secretary stated that he concurred with all of our recommendations. We also note that VA's and VBA's CIOs took quick action to address areas of concern that we've identified. As we've said in our report, we're very encouraged by these steps.
    That concludes a summary of my statement, Mr. Chairman. I'd be pleased to address any questions that you or the other members may have.
    [The prepared statement of Mr. Willemssen appears on p. 55.]

    Mr. EVERETT. Well, I thank you. I hope our D doesn't go to an I, incomplete or something.
 Page 17       PREV PAGE       TOP OF DOC
    Is it correct to characterize your overall testimony as saying that there are serious risks of significant Year 2000 failures that might affect our veterans?
    Mr. WILLEMSSEN. That's a fair characterization, but if I may add, it's also fair to add that VBA is clearly aware of these risks. They've been very responsive to the issues that we've raised, and taken action very quickly to try to resolve them, but there are remaining risks. So I think that is a fair characterization.
    Mr. EVERETT. I will add to that that I've been very pleased with the private meetings that I've had with those who are charged with this responsibility. I'm gravely concerned, though, that we are so far behind that we may not have a lot of catch-up time.
    As a matter of fact, given the very tight compliance schedule, how much margin of error do we have?
    Mr. WILLEMSSEN. It is very tight. In looking at the schedule, we believe that VBA, like other major agencies, has to set aside calendar year 1999 to perform critical testing activities to make sure that the fixes that they've put in place are actually going to work as needed.
    Related to this, we think it's very important that VBA identify priorities in what systems have to be fixed first, what second, and so on, because it may turn out that we run out of time and we can't fix everything, and we don't want to be stuck at that point with the most critical systems being the ones that aren't fixed.
    Therefore, we think it's very important that priorities be established.
    Mr. EVERETT. The VA Year 2000 strategy depends on limited financial and personnel resources. Would you say that that's a recipe for failure?
    Mr. WILLEMSSEN. Well, we've been pleased with VBA's recent change in their strategy. Actually, we think that the recent change is a less risky approach. What they were planning to do originally on many of their major systems was hope for new system developments to come on board and replace the existing systems.
 Page 18       PREV PAGE       TOP OF DOC
    Instead of that approach, they've now changed the strategy—and part of the credit goes to your subcommittee which last year push VBA to develop a contingency plan. The VBA contingency plan was to try to fix ongoing systems. That has now turned out to be their primary strategy, as they realize that we've got to make sure that ongoing systems are going to work for the Year 2000.
    Mr. EVERETT. Your report reflects the situation only at VBA.
    Mr. WILLEMSSEN. That is correct, Mr. Chairman. That's correct.
    Mr. EVERETT. Would you compare the two or where they are, or do you have any idea where VHA is?
    Mr. WILLEMSSEN. Unfortunately, we haven't done any assessments on VHA, and I don't therefore have really much of a basis to comment.
    Mr. EVERETT. Mr. Clyburn.
    Mr. CLYBURN. Thank you, Mr. Chairman. Mr. Chairman, I do not wish to take my question too far afield from what we're here for, but I—there's something that's puzzling me a little bit, and I want to ask the panel.
    If this problem is not unique to VA, and we know it's not, is there any kind of coordination taking place in the entire Federal Government through GAO or somebody to ensure that, as we bring all of government into compliance, that there is some coordination here? Is VA out doing its track and IRS doing its track, and Social Security doing its track, and everybody is doing their own thing. It seems to me that's been the failure of government.
    Is there some central place in GAO or somewhere to make sure that this whole thing is being coordinated governmentwide?
    Mr. WILLEMSSEN. OMB has set up a separate subcommittee to address the Year 2000 issue. It has representatives from each of the major departments and agencies. I or a representative from GAO also attend those meetings. They are held monthly. The next one is this Friday.
 Page 19       PREV PAGE       TOP OF DOC
    That has proved to be a good forum to share concerns, and deal with strategies to address the problem. So there is definitely sharing of information, sharing of strategies to fix the problem, because there is indeed no one magical solution, as was pointed out earlier.
    The difficulty is every agency has a wide range of heterogeneous systems, applications, database management systems, telecommunications, and operating systems. There are all different kinds and types, and they all require different solutions. So it's a big problem that has to be addressed.
    From a governmentwide perspective, OMB has had this subcommittee. I might also point out, we have several reviews, either recently completed or ongoing, at agencies where we think the risk to the American public is most severe.
    We have ongoing reviews at the Federal Aviation Administration, and Department of Defense. We've just completed this review on VBA. I testified last monthly on Medicare processing and how they were handling the Year 2000 issue. So we're trying to also target what we think are the most significant agencies in terms of the impact on the public.
    Mr. EVERETT. Would the gentleman yield?
    Mr. CLYBURN. Sure, I'd be glad to yield.
    Mr. EVERETT. His question is a very good question. One of the concerns that we have is, in fact, will DOD be able to talk to VA, and you know, there are so many different systems out there. Who is heading up this thing governmentwide?
    Mr. WILLEMSSEN. That is a major concern of ours, too. We are pushing VBA to complete as quickly as possible its understanding and inventory of all the interfaces and data exchanges it has with other entities, and also just within the department itself.
    This is a major issue. In my opinion, based on what I've seen thus far, this could turn out to be the Achilles heel within the Federal Government—the data exchange issue. To address data from external sources, agencies may have to put in some sort of bridges or filters so that data coming in from other systems that they're not sure is Year 2000 compliant doesn't come in internally to their systems and corrupt the data.
 Page 20       PREV PAGE       TOP OF DOC
    This is a thorny issue, one that has, I would say, in the last couple of months started to attract much more attention, because agencies now are realizing, well, we can fix all of ours, but what about the other guy. They have to be concerned about that.
    Mr. EVERETT. I thank the gentleman for yielding.
    Mr. CLYBURN. Thank you, Mr. Chairman. Thank you for your answer.
    I find it very interesting that you said, well, we've been looking at VBA but I really can't tell you what's happening at VHA; and if you have that problem in that field, what is happening in the broad spectrum of things.
    It seems to me in my short experience here that one of our big problems is trying to coordinate the various aspects of government, and this seems to me to be a great time, a tremendous opportunity, to put in place a system that will allow these agencies to talk to each other, at least for 100 years; and if we can get them to talk to each other for 100 years, we might be able to solve the problem ad infinitum.
    Thank you so much, Mr. Chairman. I'll yield back my time.
    Mr. EVERETT. Thank you, Jim. Mr. Mascara.
    Mr. MASCARA. I'd like to continue on with Mr. Clyburn's question about coordination. The 2000 issue is not unique to government. Does the private sector have a place at the table with the government to jointly try to solve this problem?
    Mr. WILLEMSSEN. Yes. The private sector, in particular what we've seen in the banking and financial services industries, many of those companies are out addressing this issue. In many respects, the government is going to have to rely on private contractors to come in and do many of the fixes.
    There are a wealth of tools now that are coming out from the private sector that can enable and help Federal agencies tackle this problem in a more expeditious manner. So they are there assisting and being able to support the Federal agencies.
 Page 21       PREV PAGE       TOP OF DOC
    Mr. MASCARA. Do we have sufficient resources and funding to deal with this? I heard these large numbers, billions and billions of dollars.
    Mr. WILLEMSSEN. One of the items that we feel that VBA has to identify is to clearly lay out all the costs, benefits, and risks of its Year 2000 initiatives in detail, so that it can then, from a priority setting perspective, balance those against some of its other information technology projects which are also funded.
    It may turn out that not everything can be funded, and we have to look at priorities and see what we can do in times of limited budget resources.
    Mr. MASCARA. In my former life I was an accountant, and we talked about this as we went into the 1990s, about the problem that we would expect in the Year 2000. It's difficult for me to believe that someone hasn't done something sooner to find a solution to the problem and that there seems to be some kind of a national emergency facing us when we hit the year 2000. Mr. Chairman, someone in this Government should call a summit of some sort and get the best minds possible in the world to sit down at the table to attempt to find a solution to the problem.
    I just can't believe we're sitting here saying that, as Mr. Clyburn pointed out, that you're doing something, Social Security is doing something, the IRS is doing something. I think it's too massive for any one Agency to do anything. I think everybody needs a place at the table to solve this problem.
    Thank you, Mr. Chairman.
    Mr. EVERETT. Thank you very much. Mr. Evans.
    Mr. EVANS. Thank you, Mr. Chairman. I also have some of the same feelings that my colleagues have. So let me just follow up on some of their questions.
    In the GAO's view, does VA know enough at this point to determine whether it has committed adequate personnel and resources towards fixing this problem? I understand there have been a lot of retirements that you talked about earlier.
 Page 22       PREV PAGE       TOP OF DOC
    Mr. WILLEMSSEN. They are much further along than they were even a few months ago, but they have not fully completed what we would term the assessment phase of their Year 2000 program. They do not know all the applications, especially out in the field.
    Related to that, it's not clear at this point whether they have a full handle on all the staff resources, both internal and external, that they may or may not need. That's why we've been pushing them to set priorities as part of that process.
    Mr. EVANS. I understand they have some stated timetables of their own at this point. Are those timetables not fully adequate to solve this problem?
    Mr. WILLEMSSEN. My overriding concern is that enough time is left within 1999 to engage in various testing activities. It's one thing to go in and make the fixes, but then we've got to go in and see that those fixes are actually going to work, especially when you begin integrating those systems with many other systems, both internal and external.
    Until you've tested them in a full operational and integrated environment, you won't know for sure what the results will be.
    Mr. EVANS. All right. Thank you. Thank you, Mr. Chairman.
    Mr. EVERETT. Thank you very much.
    I want to thank the GAO for the work that they've done for the past 2 years on this. I must say, I attribute this remark to no one but the Chair. Unfortunately, I have seen this progress delayed and delayed and delayed by turf battles, by a culture that, frankly, would not reach out, Mr. Mascara, to some of the best minds that we have in this country, an attitude that we can do it ourselves, when every outside expert that has looked at says, no, you cannot do it yourselves.
    We started this with the ranking member 2 years ago, and we have had a very difficult time moving forward, and I'll admit, there has been forward movement, and we're very thankful for it.
 Page 23       PREV PAGE       TOP OF DOC
    Again, I thank you for your testimony.
    Mr. WILLEMSSEN. Thank you, Mr. Chairman.
    Mr. EVERETT. I will now introduce the third panel, Mr. Mark Catlett, the VA's Assistant Secretary for Management and Chief Information Officer and the Chief Financial Officer, and ask him to introduce his staff, please.
    Mr. CATLETT. Shall I proceed, Mr. Chairman?
    Mr. EVERETT. Mark, if you would introduce your staff.
    Mr. CATLETT. Well, it's not my staff, but this is Dave Albinson, the Veterans Health Administration's Chief Information Officer; and on my right, Newell Quinton, the Veterans Benefits Administration's Chief Information Officer.
    Mr. EVERETT. Mark, let me say in the beginning, I was very encouraged at the recent meeting that we had with Mr. Gober and the willingness that all of you had to take a very serious and hard look at this problem, realizing the consequences that would develop if we could not solve it. So we look forward to your testimony here today.
    Mr. CATLETT. Thank you, Mr. Chairman.
    Mr. Chairman, members of the subcommittee, it's my pleasure to testify today on behalf of the Department of Veterans Affairs, concerning our readiness for Year 2000. We are here today to bring the subcommittee up to date on our plans and progress in resolving Year 2000 problems.
    I have submitted my full statement to the subcommittee, which I ask to be made part of the hearing record.
 Page 24       PREV PAGE       TOP OF DOC
    Mr. EVERETT. Without objection, so ordered.
    Mr. CATLETT. VA's information systems will provide uninterrupted services supporting the full range of veterans' benefits and medical care up to and beyond the year 2000. As VA's Chief Information Officer, I have established close working relationships with VA's administration level CIOs, these two gentlemen accompanying me today, to lead our effort in the coming year 2000 compliance.
    We are working vigorously to make sure VA's systems will function correctly. VA's strategic approach is to make our existing mission critical systems compliant in their current environment. We have identified our mission critical systems and assigned levels of priority to the applications supporting those systems.
    As VA's CIO, I'm responsible for overseeing and ensuring the completion of the year 2000 project for all VA systems. The VBA CIO, VHA CIO, and senior information technology managers in the National Cemetery System, and staff offices are responsible for developing specific plans and managing the projects within their respective jurisdictions.
    Both VBA and VHA have established a Year 2000 project offices that report directly to their organization's CIO. These project offices provide for the planning, guidance, oversight and technical support for their organizations' Year 2000 effort.
    On March 7, 1997, my office established a detailed internal report to track our progress in addressing Year 2000 problems. This monthly report, modeled after OMB's governmentwide Year 2000 quarterly report, measures the progress of each individual VA application.
    In addition to this formal reporting mechanism, the administration level CIOs and their Year 2000 program officials meet with me monthly to provide status reports addressing their successes and progress towards meeting their milestones.
    The monthly reports and meetings provide my office early notice, should an organization fall behind schedule. This early notice gives us the ability to work with the organizations on a corrective action to get back on schedule.
 Page 25       PREV PAGE       TOP OF DOC
    In addition to the monthly feedback I receive, my office is conducting periodic independent assessments of VA's progress in preparing for the year 2000. Last year my office conducted a Year 2000 readiness review of the major VA organizations. This subcommittee received a copy of the review in February.
    We are planning another similar independent review by a contractor during the first quarter of Fiscal year 1998.
    In managing the overall task, we have prioritized our applications using the three-tiered structure. The first tier includes systems that will directly impact the delivery of medical care and benefits to veterans or are central to the department's mission.
    The second tier includes internal agency systems used to improve timeliness and efficiency of administrative processes, operations support or producing periodic reports. These are systems whose failures would not be deemed as having a direct adverse effect on veterans.
    Finally, the third tier includes systems scheduled for discontinuation prior to the year 2000. It may include systems scheduled for elimination because there is no further legislative requirement or program need to maintain them.
    In addition to our internal activities, we are paying close attention to the services and products we obtain from outside sources. VA has been utilizing the interim Federal Acquisition Regulations' Year 2000 compliance language since it was issued in January of 1997.
    Prior to that we used the language recommended by the Federal CIO Council Subcommittee on Year 2000. The subcommittee's language was incorporated into the interim FAR.
    At this point I would like to update the subcommittee quickly on progress in each of our organizations. The information systems supporting the National Cemetery System are fully year 2000 compliant.
 Page 26       PREV PAGE       TOP OF DOC
    VBA has completed the assessment phase of its systems. VBA's plan is to complete the renovation phase by November of 1998, validation by December of 1998, and implementation by June of 1999. Every application has been addressed, and VBA has a fixed solution and a planned fixed day for all of its applications.
    VBA recently awarded four task orders to bring contractor support on board. Three of these task orders are for renovation of applications, and one is for project oversight. They also amended an existing task order to increase the level of Year 2000 support for their project managers.
    As of the end of May, 38 percent of VBA's applications are renovated and Year 2000 compliant. Another 5 percent of their applications are in testing.
    VHA completed its comprehensive Year 2000 plan on April 30, 1997. VHA's goal is to complete its assessment, including the nationwide assessment of biomedical equipment at VA medical facilities, by January of 1998.
    VHA's plan is to complete any necessary renovation by July of 1998, validation by January of 1999, and implementation by October of 1999.
    As of the end of May, 23 percent of VHA's VISTA, what we formerly called DHCP, their Veterans Health Information System and Technology Applications, are scheduled for discontinuation. In the OMB definition, these are included in the count as complaint. If we are to discontinue them, OMB considers that to be compliant.
    VA, along with other agencies, and the private health care community are consumers of biomedical equipment. The potential Year 2000 impact on biomedical equipment is a national issue, as has been noted here already today, affecting both the private sector and Federal health care communities.
    VA recommended to OMB in January that an interagency committee chaired by the Department of Health and Human Services be established to deal with this issue. The first meeting of the committee was held in May. The Food and Drug Administration, in their role as regulators of medical devices and biomedical equipment, will ensure that these devices are Year 2000 compliant.
 Page 27       PREV PAGE       TOP OF DOC
    We will coordinate a public awareness campaign with HHS as it particularly affects veterans with medical devices in their bodies or in use in their homes. Additionally, our patients will be advised as to the status of medical center equipment when the work of the HHS-led committee has identified potential problems.
    The Austin Automation Center has made excellent progress in preparing for Year 2000 as well. The AAC, as we call it, provides VA-wide information and technology support to all components within the department. As of the end of May, 74 percent of all applications they support have been renovated and are Year 2000 compliant.
    The AAC plans is to have all systems renovated by September of 1998, validated by October of 1998, and fully implemented by September of 1999.
    Mr. Chairman, the Department of Veterans Affairs is following a solid plan that will allow us to continue serving veterans and their families without interruption into the next millennium.
    This concludes my opening statement. Mr. Albinson, Mr. Quinton and I will certainly be happy to answer your questions or those of the members of the committee.
    [The prepared statement of Mr. Catlett appears on p. 71.]

    Mr. EVERETT. Thank you very much, Mr. Catlett. I appreciate your testimony. All statements, by the way, will be—entire statements, both from members and all our panels, will be entered into the record.
    I again repeat that I was very encouraged at our last meeting in my office and the candor—candid conversation that we had, and along those lines, one of the subjects that we approached—in the corporate world, if a major project fails, the managers of that project are generally fired. My question to you then and is now: If a situation of VA Y2K project—if it fails, who loses their jobs?
 Page 28       PREV PAGE       TOP OF DOC
    Mr. CATLETT. Well, Mr. Chairman, the three of us most responsible for it are sitting before you today. I know the Deputy Secretary made that same commitment that he would be personally involved and committed to that success as well.
    Mr. EVERETT. His comment was that he would be one of the first—on the train first, and the rest of you would be right behind him.
    Mr. CATLETT. I recall something very much like that. Yes, sir.
    Mr. EVERETT. Well, I want to put that in the record to emphasize how critical this committee views this situation. I come out of the private world where you either produce or you—and if you don't produce, you fail. I, frankly, cannot—I cannot imagine the business world tolerating a situation we're in here today, and I wish for you to provide for me an actual list that I can enter into the record for those who are responsible for this project.

    [The information follows:]
"The Official Committee record contains additional material here."
Insert on p. 50

    Mr. CATLETT. I'll be glad to do that, sir. I would note on your comment, this issue does exist for the private sector, and there are a lot of people facing the same pressures and same issues.
    Mr. EVERETT. Let me ask you—We also discussed this. Does the VA have adequate resources and personnel to achieve 2000 compliance? Will VA achieve total Year 2000 compliance, with the emphasis there on the word total?
    Mr. CATLETT. Sir, as we noted, we have an estimate now of $144 million. We intend to update that estimate. Our next quarterly report due to OMB is in August. The task orders and the oversight activities by contractors that I noted in my statement, particularly in VBA will assist us in updating our cost estimates. We intend to have an assessment from those contractors by August 1. It will be a portion of our updating of that estimate.
 Page 29       PREV PAGE       TOP OF DOC
    VHA will continue with their assessment that they are scheduled to complete in January. I would note that there are several issues here, and it's a question of definitions that has been discussed at the CIO Council and with OMB.
    We are replacing systems that we have not identified in that estimate, because we began to replace those systems for reasons other than Year 2000. I think the same will be true in the biomedical area.
    There's a replacement schedule for all of our equipment that we use in our hospitals. If we need to accelerate that because of Year 2000 problems, that portion we would identify as a Year 2000 cost.
    If it's scheduled to be replaced anyway in 1999, I wouldn't recommend identifying it as a Year 2000 cost, but if it's scheduled to be replaced in 2001 and we have to replace it in 1999, then it would be a Year 2000 cost.
    So we will be updating the assessment as we go, and our intention is not to low ball the estimate.
    I would add one other thing, going back to VBA. We spoke briefly in your office, as you say, a little more than a month ago, maybe 2 months ago. I'm encouraged. The subcommittee mark in the House appropriations has included an extra $7 million beyond our request in the general operating expenses account.
    It's for several issues, including the possibility of our need for more contractor support in our Veterans Benefits Administration applications. Our intention is to complete the re-coding, the renovation, as it's being called, of our existing systems by December of 1998.
    We clearly agree with Mr. Willemssen. A year of testing is required. Not that testing won't begin before that, but we want to have re-coding done, essentially completed, so that we can have a full year of testing for all the interfaces.
 Page 30       PREV PAGE       TOP OF DOC
    We've asked, and the committee has responded. We will certainly work with the Senate committee to make sure we have the flexibility, because for systems work, as you note here with your clock 1998 is an important year for us. We can't wait for a 1998 supplemental next summer.
    If we find out this August that we need funds, more funds than we've estimated in 1998, we're trying to ensure that it's included in the appropriation completed by the Hill this Summer.
    Mr. EVERETT. Finally, to ask you: You mentioned biomedical. We know there are certain safety factors, and we know that we need to assure our veterans who have pacemakers, for instance, what the situation is. My mother-in-law has a pacemaker, and she has that thing checked by the telephone, and I'm sure some of our veterans are in the same situation.
    How does the FY2K situation affect those veterans?
    Mr. CATLETT. Could I ask Mr. Albinson to provide that information for you. Thank you.
    Mr. ALBINSON. Thank you. Mr. Chairman, the VA currently supports almost 19,000 veterans with implanted pacemakers, and as the single most critical device which also has a date element involved as part of its architecture— —
    Mr. EVERETT. I understand that affects the telephone situation.
    Mr. ALBINSON. Yes, sir. We made a preliminary survey of the vendors who supply these biomedical devices to us, and I'm pleased to relay to you that, having checked 95 percent of those 19,000 devices, that we have found that they are all compliant at this point.
    Now we're going to go and continue to check that last 5 percent, but at this point I can tell you that that problem seems well in hand.
    Mr. EVERETT. And I'm sure we're getting that word out to our veterans, so that they won't worry about it.
 Page 31       PREV PAGE       TOP OF DOC
    Mr. Clyburn.
    Mr. CLYBURN. Thank you, Mr. Chairman. Let me, if I may, get back to your question, Mr. Chairman, how sure we are about compliance?
    You mentioned that by August 1997 you were going to make some assessment as to whether or not you need more resources?
    Mr. CATLETT. That's this August, in 2 months, in less than 2 months now, 5 weeks.
    Mr. CLYBURN. Do you think that in 2 months you will be able to give us a pretty adequate estimation of how much you will need to get this done?
    Mr. CATLETT. Mr. Clyburn, we have an estimate now of what our resources needs are, particularly for our systems work. We do not have that for the biomedical, but for our systems work we have an estimate now.
    We intend to utilize the contractors in the Veterans Benefits Administration area for their systems work. We intend to get their estimate as well, with the goal being to update the cost estimate to complete renovation of our current code by December of 1998, giving us the full year to test.
    So if our estimate is short, we intend to be aware of that with an August 1 report from our contractors. So we want to know their opinion, in effect, of our estimate. Recognizing that doesn't always fit with the schedule in terms of appropriation action, with the speed at which appropriation action is being completed, we are making known to the Appropriation Subcommittees, both House and Senate, that we'd like to have some flexibility in our 1998 appropriation that affects VBA so that we can provide support for our recording if the estimate is higher than what we've projected at this point.
    Mr. CLYBURN. I think you used a figure of $7 million that's been put in, an additional 7 million.
 Page 32       PREV PAGE       TOP OF DOC
    Mr. CATLETT. I just saw this this morning. That's what I understand was the very recent action. I've been on travel a few days, but sometime early this week the House Appropriations Subcommittee has included an additional $7 million beyond the President's request for the general operating expenses appropriation, which funds VBA.
    They are indicating in the report, as I understand—again, that these extra funds are for several purposes: Benefits processing improvement, as well as a need for the Year 2000 contractor support.
    Again, if in August we think we need more support than we've now projected for 1998, we're very encouraged that at this point the House subcommittee has reacted, and we clearly intend to address this issue with the Senate subcommittee as well.
    For systems work—1998 is the crunch year. Biomedical, as I see it, the crunch year is 1999.
    Mr. CLYBURN. Well, I just wanted to be sure I understood this. This is 7 million beyond what you originally thought it would cost.
    Mr. CATLETT. No, sir. The committee has added $7 million. We do not yet know if we need another million dollars beyond what we've estimated. We're going to have— —
    Mr. CLYBURN. You just asked them to add.
    Mr. CATLETT. I didn't ask for a specific amount. I just asked for flexibility. They've reached that decision, and they're looking at a lot of things; and as I said, it goes beyond just Year 2000. They're going to note, but I'm very encouraged they're giving us flexibility.
    So as I said, we have an estimate of our contractor support to complete renovation by December of 1998. We have contractors in place now that will give us their assessment as well, and we're asking for that by August 1. We'll be glad to—Obviously, we'll share that with you.
 Page 33       PREV PAGE       TOP OF DOC
    So all this is an encouraging factor, is what I'm saying to you. I wouldn't look at any one number right now, other than we do have a positive signal from the appropriations subcommittee that they recognize and agree this is an issue we need to be ready to address quickly.
    Mr. CLYBURN. Right, I agree, but I think you'll find that all of the committees and subcommittees up here are real sensitive about this, as you can imagine. I think that our colleague said it very well. We'll have 435 district offices that—that is, provided all of us are around here at that time—that will have real problems with this. So we are sensitive to it.
    I just wanted to be sure that the flexibility you're talking about, that—it would seem to me that you would have that flexibility within your request, and if this is 7 million beyond, that sweetens the flexibility quite a bit.
    Mr. CATLETT. Yes, sir. We have, as was noted, I think, by Chairman Horn, in general the approach has been that we will have to reprogram or shift our funding, instead of trying to wait for the 1999 budget, particularly in systems work.
    The work has to be well underway now and early in 1998 in order for this to be completed.
    Mr. CLYBURN. Explain something to me a little bit, if you could. I have in my notes here that $148 million cost estimate over the next 3 years. Explain to me what happens if you need additional money beyond the Year 2000. I mean, how would all this work?
    Mr. CATLETT. Well, I'm not exactly sure of your question. We will reprogram if we need additional funds, and we will update that estimate in August for you.
    Mr. CLYBURN. And that will include the entire— —
    Mr. CATLETT. It will be the VA—Yes, sir, over this period. As I noted, the biomedical estimate I don't anticipate being there by then, because the work being done and led by HHS and FDA will not be completed. I don't think we'll have sufficient responses to give you a satisfactory answer on the biomedical equipment estimate that we need by August.
 Page 34       PREV PAGE       TOP OF DOC
    Mr. CLYBURN. Oh, so that's going to be another figure?
    Mr. CATLETT. Yes, sir, because I don't think we're going to have that in 5 weeks or 6 weeks.
    Mr. CLYBURN. Okay. All right, thank you.
    Mr. EVERETT. Thank you, Jim. Mr. Evans.
    Mr. EVANS. Thank you, Mr. Chairman. Mr. Catlett, I forwarded you a series of prehearing questions, and you've raised some additional questions. I do appreciate your responses.
    Mr. CATLETT. Thank you.
    Mr. EVANS. I'd like to ask you a few of those, if I could. Your written responses indicate that not all of the interfaces for VBA and VHA related programs have been inventoried for Year 2000 compliance. The Department states that 148 out of 429 VBA interface files have been assessed, and that 57 out of 148 are compliant. Eighty-four files are not compliant, and seven files have been retired.
    In other words, the VA believes approximately one-third of its inventoried files are compliant. The Department also indicates that the VBA interface management plan is on target for completing its interface inventory by June 30, 1997, just 4 days from now.
    Since the rest of the inventory process is near completion, can you give us at least a thumbnail sketch of VBA inventory expectations? Can we expect a similar ratio of compliant versus noncompliant files?
    Mr. CATLETT. I would ask Mr. Quinton to provide that information.
    Mr. QUINTON. Yes, sir. Mr. Evans, the number indicated there, the 429 interfaces, is our latest estimate of the interfaces. The other number you referred to, 148—we showed that as the number that we had looked at to verify whether or not they were compliant, at this point in our review.
 Page 35       PREV PAGE       TOP OF DOC
    What we're trying to do is continue that process, looking at each interface and determining if there's a date issue with the interface. In some cases there may not be a date issue; in some cases, there may be. We have to look at each and every one of the group of 429 to verify whether we have a date problem or not, and that's being done system by system.
    So the total number of interfaces we determined at this point is the 429 interfaces. We will continue to look at each and every one of those by system.
    Mr. EVANS. Mr. Catlett, the VA has indicated that VHA is presently creating a profile of interfaces between its systems and other systems throughout the VA, other Federal agencies and other commercial systems. You also indicate that VHA has sent a Year 2000 compliance letter to corporate systems' owners and managers and has asked for detailed information on interfaces.
    When did this process of creating an interface profile within the VHA begin? How far along are you in the process, and when will such a profile be completed?
    Mr. CATLETT. I'll ask Mr. Albinson to provide that. I'm sorry, could you repeat the last two questions?

    (Subsequently, the Department of Veterans Affairs provided the following information:)

    An initial effort to update VHA's inventory of corporate or national systems/databases began in March 1996. One item of information requested at that time was a listing of data sources that fed each of these systems. As VHA's awareness of the Year 2000 problem increased, we decided to verify the existing information and solicit additional information on out-going interface data. The first VHA Corporate Systems Year 2000 compliance status request letters were sent to System Managers of Record (SMRs) of forty of VHA's Corporate Systems on June 2, 1997. Letters to the System Managers of Record of the remaining 129 VHA Corporate Systems were sent on July 21, 1997. These letters requested detailed information on interfaces between VHA Corporate Systems/databases and other systems, and plans for assuring that the exchange of data is Year 2000 compliant. Initial responses are requested by the end of August, 1997. Any necessary follow-up responses will be received by October 1997. The assessment phase for all Corporate Systems is scheduled to be completed by January, 1998.
 Page 36       PREV PAGE       TOP OF DOC
    We are also investigating the interfaces and dependencies of our VISTA programs with their environment. VHA will be taking steps to ensure they continue to operate correctly by monitoring their integration with other information systems within VA, other Federal agencies and commercial products or equipment. A complete inventory of external interfaces to each of the 141 VISTA applications is being conducted and a final assessment is scheduled to be completed by January 1998.
    VHA is writing to vendors of medical devices and asking them to assess their products for Year 2000 compliance. Vendors are being requested to provide a description of all interfaces; the type of data exchanged between these devices and interfaces; and plans for assuring Year 2000 compliance. VHA has mailed letters to 120 vendors of medical devices beginning on June 20, 1997. The first group of those vendors was asked to provide a written plan to VHA by July 18, 1997. While responses to date have been limited, those who have responded indicated that their equipment would either not be affected by the Year 2000, or that compliance efforts are currently underway. VHA expects to continue to send out letters as more vendors are identified.
    VHA has established an expert panel, Medical Device Integrated Product Team, to evaluate and validate these vendor responses affecting medical devices and the interfaces connected to these devices. The assessment phase for all Medical Devices is scheduled to be completed by January 1998.

    Mr. EVANS. When will the process of creating an interface profile within the VHA begin? How far along are we in the process, and when will such a profile be completed?
    Mr. ALBINSON. I can provide the actual dates that the letters went out and the expected responses for the record. I can tell you that it is part of our assessment process which will be completed by the end of this calendar year.
 Page 37       PREV PAGE       TOP OF DOC
    Mr. EVANS. Mr. Chairman, I have a number of other questions, rather technical in nature. I'd like to submit them for the record and ask that the questions and the responses be made part of the record.
    (See p. 87.)

    Mr. EVERETT. Absolutely. Matter of fact, I might add that I think the second half of my question on if the VA will achieve total Year 2000 compliance—I don't think we got around to that, and I also have some costs, how much money has been spent so far. I'll ask that you submit those for the record also.
    Mr. CATLETT. Certainly.
    (The information follows:)

Table 1

Table 2

    Mr. EVERETT. Do you have additional questions, Lane?
    At this time I'd like to recognize Dr. Snyder, also a very knowledgeable member of our subcommittee, who has joined us, for any questions he may have.
    Dr. SNYDER. Thank you, Mr. Chairman. I'm sorry I'm late. One of the advantages of being late is you can ask any off-the-wall question you want, and everybody will say this poor fellow, he just didn't attend the hearing.
    Would you help me, please, with this problem, which I know is not just a VA problem. It's a problem throughout the world with computer services. Have you all had any discussions amongst yourself looking back to the folks you contracted with five and ten and 15 and 20 years ago for your data processing, why we were set up for these kinds of problems?
 Page 38       PREV PAGE       TOP OF DOC
    I mean, it seems like the time to have corrected this was at the beginning of computerizing records, not 2 or 3 years before the turn of the century. Has there been any—You know, what's the results of some of your Monday morning, 20/20 hindsight? Have you been angry with any of your vendors, saying why didn't you all—you set us up for a problem here by not seeing this coming?
    Mr. CATLETT. Well, very generally now, I think Chairman Horn spoke earlier about it, that it was a matter of efficiency when these systems were set up to use two digits instead of four. I'm sure there was other streamlining that happened.
    It's been noted here that this is now an issue for the next 2 or 3 years; clearly, we have with the interest of the committee and others, intensified our efforts; but this issue has been addressed and identified in the VA for a decade.
    VBA began some work and made some changes long ago. Our office automation center did the same as well. So it's been underway for sometime. So again, it's—Like you say, it seems so trivial to many people. Yet it's going to cost a lot of money, and it's a worldwide problem to address.
    Dr. SNYDER. I understand that. It just seems like this is a problem that, when we purchase computer services, that we know the VA is an institution that's going to be around in the year 3000, that some of our vendors should have said, you know, what we use at a hardware store which may not be around in 3 or 5 or 10 years may not be acceptable for the Veterans Administration 20 years from now.
    Thank you, Mr. Chairman.
    Mr. CATLETT. I would make one other point. Despite the fact that we more and more go to the market to buy software, we're not a large enough share of the market to dictate that supplies change that.
    Now in order for the software to work, it has to be done; but I don't think we could have 5 years ago, even if we were more prescient than anyone else, to say, hey, you have to do this for us. We're not a big enough portion of the market to demand that type of change, I don't think.
 Page 39       PREV PAGE       TOP OF DOC
    Dr. SNYDER. I apologize for being late. Thank you, Mr. Chairman.
    Mr. EVERETT. Thank you, Dr. Snyder.
    Mark, we appreciate your appearance here today and that of your panel members. I would reemphasize how this chairman would view the failure to achieve this. It's very critical. I would add to that that it would not only, in my estimation, be the VA's failure and something ought to be done about those who fail, but it would be also this committee's failure, if we don't exercise proper oversight, if you will, to nudge VA where we feel like they ought to be nudged in solving this problem.
    I can't emphasize, as I know you realize, how great this problem is to our Nation as well as our VA.
    So I thank you for appearing here today, and if you will, tell Hersh I'll look forward to seeing him in another capacity before too long.
    Mr. CATLETT. Thank you, Mr. Chairman. Appreciate it.
    Mr. EVERETT. Thank you.
    Mr. EVERETT. At this time I would recognize our last panel, Dr. Tom Shope. He is the Acting Director of the Division of Electronics and Computer Science, the Office of Science and Technology, Center for Devices and Radiological Health of the Food and Drug Administration, and that's a lot of electronics. Doctor, welcome here.
    Mr. SHOPE. Good morning, Mr. Chairman, and members of the subcommittee.
 Page 40       PREV PAGE       TOP OF DOC
    Mr. EVERETT. Doctor, we're going to ask you to adhere to our 5-minute rule. Your complete testimony will be entered into the record.
    Mr. SHOPE. Thank you, sir. I believe you have our testimony already. I'll just briefly summarize that.
    My name is Thomas Shope. I'm the Acting Director, as you said, of the Division of Electronics and Computer Science, Office of Science and Technology, Center for Devices and Radiological Health in the Food and Drug Administration.
    I'm pleased to be here to provide information about the Year 2000 date issue and its impact on medical devices. Let me assure you that the FDA does not currently believe there will be any major impact on medical device safety from the Year 2000 problem.
    FDA is responsible for protecting the public health by helping to ensure that medical devices are safe and effective. Any computer software that meets the statutory definition of a medical device is subject to applicable FDA medical device regulations.
    An issue that has been identified as warranting review is the impact of the Year 2000 on some medical device computer systems and software applications. These products could be impacted by the Year 2000 date problem only if they use a date in their algorithms or calculations or in record keeping, and if a two-digit year format was used in their design.
    Let me explain the types of software that are used in medical devices. First, there is embedded software which is software typically contained in a microelectronic circuit or a micro-chip which controls device operation. Examples of such devices are pacemakers, infusion pumps, ventilators, and many others.
    It is very unlikely that these products would be directly impacted by the Year 2000 problem. These devices do not require knowledge of the current date to operate safely and effectively.
    For example, pacemakers do not depend on a current date system in order to function properly to support the patient. These use counters or clocks, but not explicit dates. Programmers or external controllers for these devices have the potential to be affected, but the manufacturers with whom we have discussed these issues see no problems which will not be addressed.
 Page 41       PREV PAGE       TOP OF DOC
    Second, there is non-embedded software. Non-embedded software is intended to be operated on a separate computer system, often a personal computer or a workstation. Such software devices may be used to enhance the operation of another device or devices.
    These products may also use a two-digit year format. Some examples of non-embedded software devices include radiation treatment planning systems, transmission or storage of medical images, offline analysis of EKG data—or ECG data, rather, digital analysis and graphical representation of electrocardiograph data, and systems used to adjust the rate response of pacemakers, the programming portion of using a pacemaker.
    While there is a chance that the two-digit year format may affect the performance of these software devices, we believe that the Year 2000 risk will be mitigated through aggressively working with the manufacturers, as the Center is presently doing.
    The Center is preparing to send a letter to all medical device manufacturers to ensure that manufacturers address the Year 2000 issue and review both the embedded and non-embedded software products.
    In addition, we will ask manufacturers to review any computer controlled design, production or quality control processes for their potential impact with regard to the 2-year digit format.
    The Center anticipates sending this letter very soon, and we would be glad to provide a copy to the subcommittee for the record.
    The letter will remind manufacturers that, pursuant to manufacturing regulations, they must investigate and correct devices that fail to operate according to their specifications because of inaccurate date recording and/or calculations.
    For devices that are already on the market, FDA will request that manufactures conduct hazard and safety analyses to determine whether device performance is affected.
 Page 42       PREV PAGE       TOP OF DOC
    We expect manufacturers to identify products which have a date-related problem that could affect safety or effectiveness of a device and to take the necessary action to remedy the problem.
    Again, let me stress that we do not anticipate any significant problems which would affect patient safety with individual medical devices. We want to ensure the continued safety and effectiveness of these devices by addressing the issue before it arises.
    For future medical device pre-market submissions to the agency, FDA will review design processes and features to assure that the products have been designed to perform date recording and computations properly.
    Thank you, Mr. Chairman, for the opportunity to tell you about the issue of the Year 2000 and medical devices. Let me assure you that we at FDA take this issue very seriously, as we do all problems that could affect the public health.
    We have been evaluating the possible impact on devices since early last year. We are committed to a scientifically sound regulatory environment that will provide Americans with the best medical care. FDA has looked at this issue and does not see any major problem with medical devices.
    It is the manufacturer's responsibility to meet high standards in the design, manufacture, and evaluation of their products. They are ultimately responsible for these products, but FDA will provide the regulatory framework to ensure that the collaborative effort results in the best medical device products.
    Thank you, Mr. Chairman.
    [The prepared statement of Mr. Shope appears on p. 78.]

    Mr. EVERETT. Thank you, Dr. Shope. I appreciate your testimony. As you know, there is a good amount of uncertainty out there right now.
 Page 43       PREV PAGE       TOP OF DOC
    Mr. SHOPE. Yes, sir.
    Mr. EVERETT. In your testimony, you state on page 3—are you confident that there is no safety problem? This is something on which you wouldn't want to be wrong.
    Mr. SHOPE. No, sir, we would not want to be wrong. I think the—You can't make a blanket statement that there are going to be no problems with computer controlled medical devices, but I think there will be problems of the type that are easily overseen, anticipated, and dealt with by the physicians using the devices.
    Probably the most typical kind of problem will be a date/time stamp on a record of some sort that will not be properly implemented, if that's not corrected; but we hope that, by bringing this attention and working with the manufacturers, those kinds of software upgrades can take place before the impact would occur.
    Mr. EVERETT. In other words, your testimony is that we know of no devices that would cause any serious problems to the patients at this point?
    Mr. SHOPE. I think that's a fairly accurate statement. We do know of certain devices that, if they are not fixed by the year 2000, could present problems, and we know that the manufacturers are working on those.
    An example is a radiation treatment planning system that uses a radioactive isotope as the source of the radiation for the treatment. If a computer algorithm used in that kind of a device used a two-digit date, the strength of the radioactive source might be inappropriately entered in the prescription for the radiation.
    We know that there are those kinds of problems and that the manufacturers are working on them to fix the problem. That's a software upgrade, and we expect that will occur long before the year 2000 comes around. So that's an example of a non-embedded kind of problem with a—basically, with a computer program that just needs the software to be updated.
 Page 44       PREV PAGE       TOP OF DOC
    Mr. EVERETT. Would you furnish this committee with a list of all devices, biomedical devices and otherwise, that you feel like could present a problem if they are not fixed?
    I would further ask you to follow up and, as each of these devices—as you are sure that each of these devices' problems have been solved for these devices, that you would let this committee know.
    Mr. SHOPE. Certainly, the FDA would be pleased to work with the committee and provide whatever information we can. I'm not sure that we're going to have a report from every device manufacturer as they correct the problems.
    The current regulations and the legislation do not require those kinds of reports to be presented to FDA under our current regulations. We would learn if there was a problem which presents a significant risk to health and safety, and we would be working then with the manufacturer as they correct those; but if a manufacturer has no problem, we won't necessarily know that he has no problem.
    Mr. EVERETT. And you've contacted all these manufacturers that you know of?
    Mr. SHOPE. Our letter will be going out to all 13,000 medical device manufacturers registered with FDA in the near future, yes.
    Mr. EVERETT. In June 1997 my question to you would be why the letters have not gone out earlier?
    Mr. SHOPE. We began to look at this problem last year, as I mentioned. We became aware of the need to let the device manufacturers know that they need to look at this.
    I think—and the way medical devices are developed and the software that's developed is a very structured engineering process that goes on, which is a hazard analysis to look at all the potential modes of failure and make sure that you've addressed issues to deal with those, and I think, a large majority of devices, that has been the case. The year 2000 date is not going to be a problem.
 Page 45       PREV PAGE       TOP OF DOC
    That doesn't answer the question of why we didn't do it last year. We've been doing, I guess, lots of other things, and this didn't come to the attention of us at an earlier time.
    Mr. EVERETT. I would hope it is to your attention now, and I would urge FDA to move on this as quickly as possible, and also to inform this committee when those letters have all gone out.
    In addition to that, I would like for you to keep this committee updated on what the responses are and what your evaluation of those responses might be.
    At this time, I'd like to turn to my colleague, Dr. Snyder, for any questions he may have.
    Dr. SNYDER. Thank you, Mr. Chairman. On page 5 you gave a list of the non-embedded software devices. Could we run through a few of those, and you give me some scenarios of potential problems that—why you put them on your potential list. Give me a specific example of a doctor and a patient and what could be happening in a hospital or clinic, at a VA facility that might cause—or any medical facility that might cause problems.
    Mr. SHOPE. The issue of pacemaker telemetry data, for instance—The problem that we could foresee there is likely with a personal computer type control of the evaluation of the data from the pacemaker that's received by telemetry. There it would be a date/time stamp kind of thing, you know. The date at which this is being done could be incorrect if the PC program is not recording dates appropriately.
    As you know, many personal computer type programs will have problems with the dates, if you don't take the right adjustment to deal with that.
    Dr. SNYDER. Okay. So let's give us an example. So we're getting to—It's New Year's Eve, and I've tied one on, and January 1st 00 comes along, and I go into super-ventricular rhythm or something. So tell me how then that's going to impact—How do you see that would be a problem?
 Page 46       PREV PAGE       TOP OF DOC
    Mr. SHOPE. That's not going to be a problem. The problem would be perhaps 3 weeks later when you come for your routine monitoring with the physician, and he records the data. His computer records might not have the right date associated with it.
    Dr. SNYDER. If the computer hadn't been upgraded. So— —
    Mr. SHOPE. It's not a direct impact on patient care. It's more a record keeping issue.
    Dr. SNYDER. So they try to read it, and you're saying that when he pulls off the information, because it's an 00 that it may not pull off that particular time. It might consider it the year 1900 or something and too old to be pulled up?
    Mr. SHOPE. It's not—We don't know the details of how that might work, but my assumption would be perhaps in the record it just records two digits, and you would see that this happened in 00, and you would have to realize that wasn't 1900. We didn't have computerized devices at that time. So it is 2000, but the details of those kinds of interactions— —
    Dr. SNYDER. But that's not a big problem that you just described there.
    Mr. SHOPE. Yes.
    Dr. SNYDER. It would be a big problem if it didn't pick it up, for some reason. Is that—How about offline analysis of EEG data?
    Mr. SHOPE. Well, I think offline analysis—It's again the type of—
    Dr. SNYDER. Did you mean to say EEG?
    Mr. SHOPE. No, that should have been ECG in the testimony. I think I said ECG in my oral remarks. There it would be the issue of a computer algorithm that is used to evaluate—to review the trace and provide some information to the physician.
 Page 47       PREV PAGE       TOP OF DOC
    Very often these algorithms need to know the patient's age to do that. So you might enter the birth date, and the computer would subtract the birth date from today's date and come up with an age in years or days or whatever that's used in the algorithm. So there could be an error in that calculation which would result in perhaps a misinformation to the computer algorithm that is doing the evaluation.
    That's overseen by the physician, typically. We don't rely on just these programs. So we think that's not an immediate threat to patient safety. It could lead to confusion in the patient's records. Those things do need to get fixed. It's a small computer programming type correction that will need to be made.
    It's not clear, I must say, that any of these auxiliary type programs have the problem. The manufacturer is the one who knows how he designed the algorithm. We at FDA don't have that detailed data. That's why we are saying to the manufacturers, you need to review your products, work with your customers to make any corrections that may be necessary.
    Dr. SNYDER. So let me see if I got that example right. So in that situation, we're doing an EKG in February '00, and it's going to be a computer read thing that a lot of, say, rural hospitals would use and will enter in there an age and a birth date.
    Now if the program interprets it based on the age, 95 years old or something, we'll be okay or 3 months old, but if it reads it off the date '00 and calculates it incorrectly, then your norms for a pediatric patient aged 6 months is going to be different from an adult patient in some things. That's the kind of thing you're talking about?
    Mr. SHOPE. Right, and so the determination that the computer might give in the way of advice about the trace could be in error, because it assumed the wrong age for the patient.
    Dr. SNYDER. So as more and more hospitals, I think, are moving—or clinics moving to do some telephone interpretation of things, you're going to have to know what's at the other end to be sure you're getting—or both ends are going to have to know.
 Page 48       PREV PAGE       TOP OF DOC
    Mr. SHOPE. But, ultimately, I think, even if it's a telephone interchange of information, the physician ultimately is going to look at the trace. You don't rely on the computer program solely. You can use that for confirmation, for supporting information, but I see it's not a direct impact unless somehow the doctor is not in the loop at all, and I'm not aware that— —
    Dr. SNYDER. Well, a lot of us use it like red flags. If you get a red flag back that says, oh, this is an abnormal 6-monther, and it's really because it was interpreted as being a 95-year-old, that's something.
    In terms of—You gave the example up above on the radiation. I'm one of those in my naivete that think the real problem begins at the turn of the clock January 1. In some of these kind of things, I could foresee potentially that the problem could actually present itself before we get to the year 2000 if you're looking ahead. Okay?
    It's, you know, 1999, February, and we're going to be calculating your dose and over the next year to get the '00, that somehow it could ensnare problems prior to the year 2000. Am I off base on that? As a computer programmer would look ahead over the next, okay, 18 month course of therapy, and then we enter an '00 for treatment number 17— —
    Mr. SHOPE. Yeah. I'm really not expert in the therapy delivery, but I'm not aware that there are very many regimes that spread it out over that length of time. It's typically a few weeks or a few months.
    The issue is going to be— —
    Dr. SNYDER. I see what you're saying.
    Mr. SHOPE. The issue is going to be in any of those situations how the algorithm has done the two-digit date, and if it's simply it works okay with 1999, but doesn't work okay with double zero, that's not going to be a problem until they plan a treatment after the year 2000. It's the planning and not the delivery that's the issue here.
 Page 49       PREV PAGE       TOP OF DOC
    Dr. SNYDER. Mr. Chairman, may I ask one final question?
    Mr. EVERETT. Certainly.
    Dr. SNYDER. Just a quick question unrelated to your role here, but in terms of the manufacturers and your relationships with them, do we not also have at play in this whether—I'm a family practice doctor—whether we like it or not, the whole medical malpractice, the liability considerations?
    I mean, these manufacturers, if you don't do anything, I mean they have a responsibility. If they get no notice from you, I mean, they need to be looking ahead at these kind of problems and have an absolute responsibility to get this clarified and cleaned up, so that family doctors like me can rely on this stuff. Is that a fair statement?
    Mr. SHOPE. Yes. I think manufacturers, in order to have satisfied customers who expect the products they've bought to meet the specifications and the function—Nobody that I'm aware of contracted to buy devices that wouldn't work after January 1. So I think there's a very large incentive for manufacturers to satisfy their customers' needs here.
    Dr. SNYDER. I mean, hopefully, when you send out your 13,000 letters, you're going to get 13,000 responses that say we've been looking at this for 5 years, and it's all taken care of.
    Thank you, Mr. Chairman.
    Mr. EVERETT. Thank you. Dr. Shope, I must tell you, I'm extremely disappointed in the FDA, the fact that they have not sent these letters out yet. This chairman had hearings on this almost 2 1/2 years ago, and in my opinion, it was at that time too late to start some of the actions that we need to start such as the biomedical.
    Very frankly, I can't imagine that FDA would sit over there that long and still not have had these letters out. As I said, this chairman had hearings in Pensions and Compensation Subcommittee almost 2 1/2 years ago on this problem, and the problem was known prior to that.
 Page 50       PREV PAGE       TOP OF DOC
    I would hope, as I said, that these letters go out immediately. I would also point out that I have had very candid conversations with our friends over at VBA. This is a problem of the most serious nature, not only to the veterans but to the Nation as a whole.
    I don't enjoy putting pressure on people, but as I said earlier, I view this as a joint failure, if this is not achieved. Now I do know that the folks over at VBA depend on information coming in from you, and the process of giving that information to them hasn't even started yet, the result of those letters.
    So I would again hope that VHA is given the information that they need as quickly as possible. I do thank you for your testimony here. We will have additional questions that we will put to you for the record.
    Mr. SHOPE. Yes, sir. I'm sure we would be glad to provide those answers.
    Mr. EVERETT. Thank you very much.
    In closing, let me say that, based on today's testimony, I must say that I am extremely concerned about the prospect of major Year 2000 computer system failures, but I'm also happy to hear that the Veterans Benefits Administration has heeded the GAO's recommendations and has taken quick action to address the areas that they have identified.
    The subcommittee will continue to closely follow VA's efforts to ensure that their computer systems will be able to provide uninterrupted benefits and safe and quality health care to our Nation's veterans.
    The subcommittee will hold follow-up hearings to review the VA's progress, as I said. Finally, I expect the VA to immediately inform the subcommittee of any missed milestones in their compliance program.
    This hearing is closed. Thank you all for attending.
    [Whereupon, at 11:10 a.m., the subcommittee was adjourned.]
 Page 51       PREV PAGE       TOP OF DOC