Ready, Set, Go-Live: Assessing VA’s EHR Modernization Deployment Readiness
No chapters available yet
Legend
Transcript markers
- text Sentence being played
- text Search result
- text Intervention cited by the link you opened
Click a sentence to play the video from that moment. Select text to quote it with the speaker, timestamp and link.
More options
House Veterans' Affairs Subcommittee on Technology Modernization U.S. House of Representatives
Follow its sittingsSpeaker 2
They expect safe and timely care, and we all expect systems that support our doctors, not work against them. Technology should be a tool that opens doors, not a barrier that adds more steps, more clicks, and more frustration. We heard about some of that earlier in this committee term when we saw that providers were getting frustrated with some of the interfacing with the system they have. When we first met on this topic in February, VA was just emerging from a very long pause. While progress has been made since then, we know that significant work remains before we go live.
VA has standardized over 1,000 workflows into a national baseline, with Michigan being the first to use it. VA has tightened their timeline. And for the first time since the pause, we saw large system updates roll out this August without disrupting care. These are meaningful, good signs, but we cannot ignore other red flags that are warnings. Behind the scenes, many tools slated for Michigan have never been tested on a large scale.
34 new complex clinical workflows will debut there for the first time. VA plans to test across four sites simultaneously, a strategy that leaves really no margin for error, and something that I have concerns about the risk associated with that. We need assurance that this plan is feasible in the real world, not just on paper or in a computer lab. We need to know that the lessons of the past have been learned, and not just observed and acknowledged. The user experience also remains a concern.
While satisfaction is slightly up, more attention is needed. The committee has heard from physicians that the critical function system remains unstable. We hear from VA pharmacists. I know that's a very unique role that the VA has, that tools for monitoring drug interactions are still a major pain point. VA staff are now burdened with more manual processes to ensure patient safety with drug interactions.
We need to know where these issues stand today, so Michigan clinicians and veterans are not left holding the bag on day one. And finally, we must address the sheer scale of the costs. This program began in 2018 with a $10,000,000,000 price tag. It quickly ballooned far beyond its original expectations, and the latest estimate stands at $37,000,000,000. We cannot keep writing blank checks that risk taxpayer money and slow down or worse endanger delivery of veteran care. I'm encouraged by the momentum we've seen.
I'm encouraged by the commitment of my friend Secretary Collins and the Trump administration team, but encouragement only goes so far as reality sets in. We need proof, and we need transparency. The clock is ticking down in Michigan for this to go live, the time for promises is over. The only acceptable result is a flawless go live because our veterans cannot accept failure. So thank you again for being here today.
And I do wanna say it's never my intention to take gratuitous shots at anybody appearing before this committee. But we do have a role to play in oversight of what's taking place, and we intend to vigorously and robustly carry that forward. And with that, I will yield to ranking member Buzinski for her opening statement.
Speaker 4
Thank you, mister chairman, and I agree with you. And thank you to our witnesses for being here today. It seems we are bookending this year with EHRM hearings, and I look forward to hearing how the program has improved and how the department has prepared to resume go-live activities. But to start, I do wanna address an article that I read over the weekend about Secretary Collins' plans to eliminate 35,000 physicians from the Veterans Health Administration. I am disappointed that we had to learn about this through the media and not from the secretary himself.
Something this significant, I believe, warrants proactive communication with members of Congress. The department is quick to say that most of these positions are vacant, but the word is most, and they don't use them all. The VA workforce is already stretched too thin. Eliminating positions does nothing to help veterans who, in many places like my district, are waiting months for appointments. In fact, these actions threaten to undermine VA's ability to deliver timely care.
Only at the VA can veterans expect to receive care from providers who have a deep understanding of their unique experiences. That's irreplaceable. If Secretary Collins is serious about keeping veterans at the center of everything, VA health care must continue to lead the way. VA must be fully staffed and resourced. The continued efforts of this administration to bleed the VA dry will make it harder for us to honor the service of our nation's veterans.
Their efforts to outsource veterans' care will also have dire impacts on the success of the EHRM program. Today we are 117 days from VA's EHRM go live at four VA medical centers in Michigan. After the program was reset for three years, the department made the decision not only to restart, but to accelerate the go-lives in an effort to finish almost on schedule. I am sure the chairman is anxious about this. I know I am.
There are VA facilities around my district that are scheduled to go live right after the Michigan sites. The veterans of Michigan's 7th District and Illinois' 13th District are next to be impacted by EHRM. We need assurances that Oracle and VA have fixed the issues that are still plaguing the first 6 sites. At the time of our last hearing on EHRM in February, the department had dozens of outstanding recommendations from VA's Office of the Inspector General and the Government Accountability Office. According to GAO's testimony, it hasn't changed.
Our goal is to ensure that we are setting VA up for success. However, what I have heard in the past year has not convinced me that VA is ready for launch at 13 facilities in 2024. I have raised many questions with VA and Oracle, but the answers do not give me confidence. In fact, I worry that we are spending billions of dollars while simultaneously setting this program, particularly the six sites that are already live, up for failure. I need both VA and Oracle to tell me what they have done to address concerns raised by VA employees and veterans at the first six sites, such as prescription errors and incorrect alerts.
For veteran patients, the consequences of these errors range from discomfort to death. We must know that catastrophic errors in the system are not putting veterans' lives at risk. We must know that catastrophic errors in the system are putting veterans' lives at risk. I also want to hear what state the system is going to be in when EHRM goes live at the next 4 sites. Earlier this year, we learned the VA was looking for an EHRM systems integrator.
I was hopeful that this might be a positive change in the program rollout. There are a lot of questions about this contract, which was awarded to Accenture in November, and how it is going to work. In fact, ranking members Takano and Blumenthal sent a letter to Secretary Collins shortly after the award was announced and have yet to receive a response. What will Accenture actually do to make EHRM more successful? Is it a true systems integrator or is it a continuation of the role currently held by Booz Allen Hamilton?
How will Accenture be effective given its lack of authority over the Oracle Prime contract? Unfortunately, the more I hear about this contract, the more it seems to be a replacement for VA's program management contract, not an actual systems integrator. September, deputy secretary Lawrence informed me that VA estimates the lifecycle cost of this program to be $37,200,000,000, but it seems like this estimate has changed quite a few times. VA originally told Congress, as Chairman Barrett mentioned, they needed $10,000,000,000 for the entire program. Shortly after that, they came back and asked for an additional $6,000,000,000 for VA's program cost.
In 2022, VA contracted with the Institute for Defense Analysis to conduct a true lifecycle cost estimate, which shifted the estimate for the project to almost $50,000,000,000. And now the deputy secretary is backtracking and saying that it's only going to cost $37,200,000,000. I'm concerned that nobody actually knows what the bottom-line cost is. I need to hear today that VA has a grasp on this. The American taxpayers and veterans deserve transparency.
Finally, I want to address a recent article in the Washington Post about the VA's EHRM program. I believe this article highlighted real concerns from department employees about the system, concerns expressed from a place of worry for patient safety and provider burnout. I wish the department would take these concerns seriously, and use that feedback in their change management efforts. I find it very troubling that the department instead seems to be minimizing the concerns raised. Their framing of the story takes their usual tact: blame everything on the previous administration.
And I'll be honest, I think there is plenty of blame to go around. But, of course, it was the first Trump administration that rushed the VA into a sole source contract with Cerner before either was ready. Ultimately, we should not be pointing fingers. We need to have the difficult conversations to make sure that both Oracle and VA are accountable to Congress, to VA employees, and, most importantly, to veterans. We need to ensure that this new EHR supports VA's provision of world-class health care.
Thank you, Mr. Chairman, and I yield back.
Speaker 2
Thank you, ranking member Buzinski. And I will now introduce our witnesses. From the Department of Veteran Affairs, we have Doctor Neal Evans, acting program executive director of the electronic health record modernization integration office. Did I say that correctly, doctor? Alright.
Very good. From Oracle, we have the honorable Seema Verma, executive vice president and general manager of Oracle Health and Life Sciences. And from the GAO, we have Ms. Carol Harris, a familiar face to this committee, the director of IT and cybersecurity at GAO. I ask the witnesses to please stand and raise your right hand. Well, your right hand.
Do you solemnly swear under penalty of perjury that the testimony you're about to provide is the truth, the whole truth, and nothing but the truth? Thank you, let the record reflect that all witnesses have answered in the affirmative. Doctor Evans, you're now recognized for five minutes to deliver your opening statement on behalf of the VA.
Speaker 5
Thank you. Good afternoon, Chairman Barrett, Ranking Member Brzezinski, and distinguished members of the Committee, including Mr. Luttrell. I want to begin by thanking Congress and this committee for the opportunity to testify today, and for your continued support of VA's electronic health record modernization efforts. VA remains committed to successfully implementing a modern, interoperable electronic health record system, which we refer to as the federal EHR.
And we intend to implement that across the entire VA enterprise. As was mentioned since our last hearing in February, VA has made significant progress toward meeting that goal. In March, VA announced its plans to deploy the federal EHR to nine additional VA medical centers and associated clinics in Ohio, Indiana, Kentucky, and Alaska by the end of calendar year 2026. Those were in addition to the four previously announced medical centers in Michigan slated for deployment, bringing the total number of planned deployment sites in 2026 to 13. That encompasses more than 100 physical locations when clinics are considered, in addition to the medical centers, and will involve transitioning more than 27,000 VA employees from Vista to the federal EHR in this coming year.
In addition, we've begun deployment work at seven additional facilities, with 19 more on the way, all with planned go-lives in 2027. Furthermore, as evidence of our commitment to full implementation of the system across the VA enterprise, we recently shared a schedule with this committee outlining our plans to complete deployments of the federal EHR at all VA medical facilities as early as 2023. Based on lessons learned, VA will now be using a market-based approach for deployments with multiple medical centers working together and going live simultaneously in each deployment wave. This approach allows us to scale up the number of deployments, enhance efficiencies, and improve the sharing of best practices within and between markets. Now I'd like to bring our focus back to the near term.
As was mentioned, we are only 117 days, less than four months away from our planned go lives in Michigan, and less than a year away from the remainder of our deployments across all of VSN-ten in Alaska. The title that you chose for this hearing is apt: ready, set, go live. Teams at all the 26 sites have been preparing diligently for upwards of a year. The gun has already gone off. Go live dates are imminent, and we are on track for successful deployments.
Contributing significantly to VA's current momentum has been leadership ownership of this project at all levels of our organization, starting at the very top. Since assuming their roles, Secretary Collins and Deputy Secretary Lawrence have prioritized attention to this critical project and have taken significant steps to ensure that all stakeholders are aligned. Doctor Lawrence has been talking to site and vision leaders weekly and has made multiple visits to sites scheduled for go live in 2026, where he has made it a priority to listen to fellow veterans and VA staff talk about their experiences and expectations, and to respond quickly when adjustments are needed. Just as important, leadership and staff at both the individual sites and the division level are encouraged by the program's direction and newfound momentum, and are perhaps the most important drivers behind our current progress. As for the system itself, VA and Oracle Health have made significant strides in assuring the federal EHR is performing reliably and meeting our expectations.
As an example, as of November 19, Oracle Health Systems within the federal EHR maintained an incident-free time of 95.93%, exceeding our agreement of greater than 95% for 21 consecutive months. We're also delivering system enhancements, new features, and system changes that are responsive to the feedback we're hearing from VA staff and are driving improved standardization across the enterprise. For example, I know we've often spoken about the system's pharmacy capabilities in this committee, and we'll be delivering 7 further pharmacy improvements before our go-lives in Michigan in April. And as another example, we just released new system functionality called Seamless Exchange in September, reducing by more than 95% the volume of external data requiring manual review and reconciliation by VA clinicians. We are seeing evidence of the results of the change we have made.
We continue to administer the federal EHR user experience survey twice yearly, and I'm pleased that we have seen consistent improvement survey over survey. Ultimately, our goal is to deliver an EHR system that earns the trust of veterans, clinicians, and staff. And that means a system that works efficiently, enhances care coordination, reduces administrative burden, and improves health outcomes for veterans. With the partnership of this subcommittee, we look forward to continued and accelerated progress over the remainder of the 119th Congress and beyond.
Speaker 2
Thank you, Doctor Evans. The written statement of Doctor Evans will be entered into the hearing record. Miss Verma, you are now recognized for five minutes. We'll review your opening statement on behalf of Oracle.
Speaker 6
Good afternoon, Chairman Barrett, Ranking Member Buzinski, and members of the subcommittee. Thank you for the opportunity to speak with you today about Oracle's work with the VA's EHR modernization program. Since the last hearing, VA announced an accelerated deployment schedule that will complete the full implementation at all sites by 2031. Next year, the federal EHR will go live at 13 sites, and we are preparing to launch at 26 facilities in 2027. I am pleased to report significant progress on the accelerated deployment plan, and that Oracle is confident that we're prepared and fully aligned with the VA to meet this goal.
Most importantly, leadership engagement and ownership of the implementation has never been stronger. There is clear direction from the highest levels and a deep sense of accountability across all partners. We also want to recognize and commend VA leadership for their clear commitment to standardization across the enterprise. Their decisive leadership has strengthened the program and demonstrated a shared determination to deliver a unified, high-performing EHR for veterans. Overall, system performance is strong.
The system is stable, and there is a notable decrease in interruptions to end users. We have met or exceeded the 95% incident-free time requirement for 21 consecutive months, and we have been free of any system-wide outages for eight consecutive months. These improvements are a direct result of the coordinated effort between Oracle and the VA. Together, we have formalized our incident review process and implemented proactive monitoring and targeted upgrades. We have optimized the EHR to improve veteran health outcomes, enhance provider productivity, and strengthen financial performance results.
With these optimizations, productivity has increased at all facilities, and cash collections exceeded FY '25 goals, achieving 180% of target. Lastly, as Doctor. Evans referenced, we have expanded seamless exchange to all federal EHR facilities. Seamless exchange compiles and deduplicates patient data from multiple sources. This reduces the volume of external data that requires manual review, and enables improved medical charting and decision-making efficiency.
Finally, we have strengthened every part of our deployment methodology, with a focus on improving staff readiness. This includes enhanced training and change management, both virtual and in person, that allows end users to not just learn about changes, but to actually try the system out in advance of go-lives. We are modernizing the underlying infrastructure through the migration of the federal EHR to Oracle's cloud. This move will not only provide better system performance and security, but will also allow VA to adopt modern tools, including our clinical AI agent. The clinical AI agent reduces clinical burden and supports safer, more efficient care, all while ensuring the provider remains at the center of decision-making.
Our work does not end here. With VA's commitment to a commercial solution, they will continue to benefit from Oracle's ongoing innovation, including our new voice-first ambulatory EHR recently certified by HHS. With this certification, ambulatory clinics across the United States, including the VA, can begin planning for the adoption of our transformative EHR. Unlike other EHRs, Oracle's was built from the ground up on a secure, modern cloud. This allows for streamlined clinical workflows and automation of manual tasks, so providers can spend more time with patients.
Oracle continues to lead on data interoperability. We have taken the White House Interoperability Pledge and have recently been designated as a Qualified Health Information Network, or QHIN, by HHS. This achievement will enable broader and more secure exchange of patient data records across disparate systems, as veterans receive care from multiple sites and multiple providers. Our QHIN will streamline connectivity, enhance data accessibility, and help ensure that every care provider has timely, comprehensive information to support better outcomes for veterans everywhere. In closing, as we move into 2024, we are confident and prepared to deploy the federal EHR under the accelerated deployment schedule.
This will bring a unified health record to veterans, which we can all agree brings incredible opportunity to improve the experience with the VA, and the health of our veterans. Thank you, and I look forward to answering your questions.
Speaker 2
Thank you, Miss Verma. The written statement of Miss Verma will be entered into the hearing record. Miss Harris, you're now recognized for five minutes. Deliver your opening statement on behalf of the GAO.
Speaker 7
Thank you. Chairman Barrett, ranking member Bozinski, congressman Luttrell, thank you for inviting us to discuss the readiness of VA's EHRM program. In June 2017, the department initiated this program to replace the legacy Vista system and has since deployed the new EHR to six of its medical centers at a cost of about $12,700,000,000. The rollout of the system has been met with poor user satisfaction, change management issues, and slow resolution of trouble tickets, other things, given the magnitude of user concerns via paused deployments in April 2023 to improve the system and address those concerns. In December 2024, VA announced it would resume deployments starting with four Michigan sites in April. By 2026, VA plans to complete another nine sites under an accelerated deployment schedule and roughly 170 more throughout the nation by 2031.
Over the past five years, we've issued 5 reports on VA's efforts to deploy its new EHR system. These reports describe actions taken by the department and identify challenges with key planning tools critical for program oversight. We also reported on challenges experienced with the initial deployment, such as the ones I just noted. To address these challenges, we made 18 recommendations, 12 of which we have marked as priority because of the critical impact they have on strengthening successful future deployments. While VA has taken action to address our recommendations, it has not fully implemented 16 of them.
I'll highlight the 12 priority recommendations here. The first two are that VA needs to produce an updated cost estimate and schedule. The latest independent cost estimate of roughly $50,000,000,000 does not reflect the many changes and delays to the program. I know you've said that they have provided you with an updated estimate of $37,000,000,000. We have not received that estimate. And so I would ask the department to provide that to us, so we can review it.
But the updated estimate is imperative to understanding the full magnitude of VA's investment. We have also yet to see an updated integrated master schedule. Consequently, as the department increases its momentum to complete 170 total site deployments by 2031, more information critical to controlling risks and informing congressional oversight is needed. Additionally, more work is needed to demonstrate the results of VA's actions to address user concerns and system issues. In May 2023, we reported on gaps in VA's organizational change management activities for EHRM.
We also reported that users expressed great dissatisfaction with the new system, and that VA did not adequately identify and address those issues. We made 10 priority recommendations to address change management, user satisfaction, system trouble tickets, and independent operational assessment deficiencies. VA concurred with those recommendations. And as of December 2024, VA has partially implemented one of the 10 priority recommendations and continues to work towards implementing the remaining nine. For example, VA partially implemented the recommendation to address users' barriers to change.
To do so, VA developed plans to address user concerns about the new system, identified in a strategic review of the program. However, VA has not yet adequately demonstrated that corresponding improvement projects have fully addressed the underlying barriers. VA has also not yet approved and implemented a VA-specific change management strategy to formalize how it will improve the readiness of end users to adapt to working in the new EHR system. Further, VA has no plans to conduct an independent operational assessment or an IOT&E test to determine whether the system is operationally suitable. Without an IVNV, the department increases the risk of deploying the system prematurely, thereby posing unnecessary risks to patient health and safety.
Moving forward, it will be critical for VA to address the 12 priority recommendations, along with the other four open ones, as soon as possible. Until these are fully implemented, future deployments are at risk of prolonging challenges like those experienced in the initial deployments. Doing so will position VA to more effectively deliver a modern health record system our veterans deserve. And that concludes my statement. I look forward to addressing your questions.
Speaker 2
Thank you, Miss Harris. And the written statement of Miss Harris will also be entered into the hearing record. We'll now proceed to questioning. I'll recognize myself for 5 minutes. Miss Harris, thank you.
You mentioned some of the cost-associated considerations and concerns that are out there, what the true price tag of this is going to be. Doctor Evans, is this something that you can furnish through the department over to the GAO for them to review, that lifecycle cost analysis?
Speaker 5
Yes. We provided that to this committee on September 30, and we certainly can provide it to the GAO.
Speaker 2
Okay. Appreciate that. And then, anecdotally, I go to the predominantly Battle Creek VA hospital near my district for most of my care, occasionally go to the Lansing outpatient clinic there. And I'll usually ask the folks that I interact with kind of their thoughts on this coming up. They're all familiar with it.
I mean, there are posters up. You can't pull into the driveway of the hospital without seeing this announcement of the, you know, unrolling of a new electronic health record management system. And, I don't advertise to them the, you know, role that I have on this committee of oversight of this process. I'll just kinda get their opinion as to how it's going. And I would say it's not great.
There's a bit of a mixed opinion. Some of them feel like they're not fully ready for the new system at this point. Doctor Evans, can you walk me through, I mean, I understand we're not gonna train people on this prematurely, but do you have confidence as the department of confidence that the end user, the people that are actually interfacing with this program, are going to be well equipped to deal with it at the time that we go live?
Speaker 5
Behind that answer. In for the four sites in Michigan, we just completed last month the super user training. So superusers are folks who are in every department of the medical centers in Michigan who will be the kind of experts in the federal EHR for their peers. I think there was over 400 folks who went through superuser training. I don't have the exact number here, but it's got a sizable number.
They are the superusers gonna basically help the end users when they start their training. For the average user, so somebody who has not been selected as a super user, they start their training February 1, so February. With superuser training, we have introduced a lot of changes to our training and how we support adoption for the Michigan Go lives. We had 96 of those we expected to complete superuser training. On average, they rated the classes 4 out of 5.
We had a lot of direct feedback that this was significantly better than the first time that for those who had gone through this before, we paused at the Michigan sites, they said, yes. This has been much better. And in addition, for the super users we've added, and for all end users, something called learning labs. This is a success at FHCC. Learning labs are when we finish classroom training or what's delivered virtually, virtually instructor led training.
After that, we do a learning lab, which is where users can come together and practice using the system in the sandbox together with their colleagues and see exactly how it works. We had 13 scenarios at FHCC in North Chicago. We went live there, got a lot of great feedback about that. We've now built 98 scenarios, and went through those learning labs with the superuser, got a lot of great feedback about it. So I would say that the end users should expect their confidence to start to grow as they get into training in February, and learning labs in March.
Speaker 2
Okay. Thank you. And then the, miss Harris, the IV and V test that you pointed out, is that the one, I think you had mentioned to me separately that the kind of four parallel testing going on instead of sequential testing, is that part of the IV and V, or is that a separate test that would be required?
Speaker 7
Well, given the change in the strategy for testing with this market-based approach where they're gonna be basically simultaneously testing at four sites, it will make it more difficult to do an IVNV, which is that end-to-end test.
Speaker 2
And both are, in your opinion, complicating, you know, could lead to a significant risk?
Speaker 7
I do believe so. I think that the simultaneous testing at the four sites, it will take a tremendous amount of resources. And to deal with the issues that come up inevitably with a go live, to be able to handle it at all four sites simultaneously could be, you know, significantly risky for the department.
Speaker 2
Miss Verma, when I get back to my next order of questions, I'll have a question for you about that. But beforehand, I wanna you'll drink your member Buzinski for five minutes for her questions.
Speaker 4
Thank you, Mr. Chairman. Doctor Evans, VA's plan to accelerate EHR deployments. We'll rely on having sufficient manpower. In the history of this program, the program office has never really been fully staffed. So I have a couple of quick questions.
How many open positions do you currently have in the program office?
Speaker 5
It's just over 100, but that's also because we just had a new signed org chart, which increased the number of positions that we are authorized to hire. So we are now actively hiring to fill the additional positions that we've added to the program office docket.
Speaker 4
So your plan is to fill all of those positions, then? Absolutely. Okay. And how will your current staff levels be able to not only resume go lives but support the four go lives at one time. But support up to four go lives at one time.
I guess that would be taken into account the 100 that you would be adding.
Speaker 5
Yes. We'll be, we are actively hiring. In addition, at the sites that are going live, there's active hiring going on. There are 510 positions in the recruitment. Actually, I think 163 of those folks are already on board at the sites that are going live to support operations locally.
Plus, we're hiring in the program office. And then, of course, we have a significant amount of help from our contract partners, Oracle, Booz Allen, and Accenture Federal Services, as you just heard.
Speaker 4
Okay. So you have significant hiring that needs to happen, though. There are a lot of vacancies in the program office.
Speaker 5
There are vacancies in the program office in part because we are expanding the size of the program office. Mhmm. From the perspective of our ability to execute the go lives in April in Michigan, we are confident that our current staffing is sufficient to get us there.
Speaker 4
Okay. Miss Verma, the scope of what Oracle and VA are planning to do is almost unprecedented. So the number of simultaneous go-lives and the time between the waves will require a massive pool of contractors to support all of these facilities, as Doctor Evans has mentioned. How is Oracle going to ensure that these people understand the system and the VA well enough to be helpful?
Speaker 6
Sure. So first I would say that Oracle does implementations all over the world for systems. So, you know, this amount and the number of sites is not unusual for Oracle as a worldwide company in terms of the deployments. That being said, we also continue to add more staff to our teams to make sure that we can scale with the deployments as well. But I think this is not an unusual thing for our company, and so we feel very well prepared to deal with the expansions and the challenges of something of this scale.
Speaker 4
But can I just drill down just a little bit more beyond just having the bodies, the VA itself is unique in its mission and culture? So when someone were to hit the ground, what is Oracle doing to make sure that additional capacity understands the uniqueness of the VA and its challenges?
Speaker 6
Sure. Well, there are requirements around training, then there's also some federal certifications that are required as well. So it's not like we interview somebody and put them on the ground. There's some training that happens internally. And then there's also, like I said, the federal certifications and some required training that the federal government requires as well around security, as well as the specific needs around the VA.
Okay.
Speaker 4
Doctor Evans, as I mentioned in my opening, I have serious concerns about the secretary's plan to eliminate 35,000 positions at VHA. In a department that has been chronically understaffed, where veterans sometimes wait months for appointments, how will these cuts impact your program?
Speaker 5
They won't. As I just mentioned, we are hiring additional staff at the sites where we are going live with the federal EHR over the course of calendar year 2026, 5,10 additional staff. Currently, recruitment is ongoing. I don't anticipate any issues.
Speaker 4
Okay. And then just to go to the lifecycle cost estimate questions, does the VA have a definition for lifecycle cost estimate, Dr. Evans?
Speaker 5
What we provided to the committee was what we call a program cost estimate. It is the estimated cost for us to complete deployment of the federal EHR across the enterprise by 2031, and to operate that federal EHR, that is to support, or what some might call sustainment, to support that EHR at the existing sites that have gone live. And that cost estimate, that program cost estimate includes all money spent in the EHR appropriation to date, and our estimate of what it will take to get to the finish line of finalizing deployment. Again, not just finalizing deployment, but finalizing deployment and supporting the operations across.
Speaker 4
Can I just interrupt? Does that break down the program versus contract expenditures, like the details? Or is it just a top-line number?
Speaker 5
It does. It does break down, and we broke it down into 4 categories. One is the implementation costs. So that's, you can think of that as the cost to actually deploy the system. The second category is site and system operations.
That in many ways is the cost of running the system, the hosting of the system, help desk support, and the operational support. The third being infrastructure. I think we've talked about this in this committee before, that a significant part of the spend here has been an uplift of the IT infrastructure to support the new modernized EHR that has sort of had to occur in parallel with the EHR rollout. And then the fourth, we call office operations, but it is really the staff, both government staff and contract staff, necessary to deploy the EHR.
Speaker 8
Thank you, Mr. Chairman. Doctor Evans, what's the dollar sign on sustainment for this software once EHR is implemented in all 170 sites and running?
Speaker 5
I don't have a number for you that is specific to the sustainment. What I can tell you is that our estimate in the final year, when it comes to system operations, which
Speaker 8
I mentioned after 2031 moving forward, what are we gonna have to pay for this?
Speaker 5
Yeah. I would estimate it's the our again, our estimate for the operations in the final year that we estimated was 2,100,000,000. Annually? Correct.
Speaker 8
Miss Vernon, if we have four sites, we're gonna go live in April. I'm sure Oracle's hovering over the top of those four sites. If those four sites, going off of what Miss Harris laid out for us, if those four sites fail, are there other sites we will continue to move forward and fire up those other sites or is it an all stop evolution?
Speaker 8
Know? So Oracle is not doing contingency planning on if these sites fail a hundred and seventeen days?
Speaker 6
Yeah. So first off, I would say that we are focused on a successful deployment.
Speaker 6
Everything we can to make sure that happens, training and support, testing of the system, making sure that the teams are adequately supported. You heard about.
Speaker 8
So I yes, ma'am. I got you. I'm just the worst case scenario. We, I'm that's the kind of guy I am. Worst case scenario, these four sites don't go like they should.
What is the contingency plan for that day?
Speaker 6
Well, you know, when something goes live and there is a deployment, there are teams in place. There's elbow-to-elbow support. We have war rooms. So if there's an issue that's going on, we are rapidly able to assess what the problem is and fix it. The other thing is that over the last few years, we've been doing a number of optimization projects.
Right? So you heard about our capability block updates.
Speaker 8
Yes, ma'am. But if those four sites fail, is Oracle gonna continue to move on the other sites that need to be activated, or is this it?
Speaker 6
Yeah. That's a decision, I think, that we would discuss with the VA and figure out the appropriate course. But, like, I.
Speaker 6
We have not had that discussion. Our discussions are focused on successful implementations. And because of the previous experiences, right, we've gone through a number of deployments. There have been a lot of lessons learned. If we look at the past deployment that we had at Chicago, again, that went very well, and we continue to learn.
We continue to do upgrades. And so far, in these last few upgrades, as well as in Chicago, those have gone successfully well. So we have no reason to believe that there would be a total failure of the system because we haven't seen that in our, you know, the last few projects, whether it's our optimization projects or the go-live at.
Speaker 8
Shipper. You know, I hope it doesn't happen. We've been waiting 10 years for this thing to work correctly. Just from my experiences in the past, if you're not contingency planning on a worst case scenario when it shows up, we're in a lot of trouble. Fair enough.
How are the sites chosen? We have these Ford in Michigan that are being activated. We had Chicago earlier. We have six sites online if I'm reading this correctly. How are the specific sites?
Do they is there connectivity currently between sites that we're going live on, so it's an easier lift. And then the sites that are kind of expanded out in, like, where I live, they're gonna be the last ones to get a shout at it, mister Evans.
Speaker 5
Yes. I can answer that. And let me just, I will answer that question about how the sites are selected. I wanna quickly correct the record on your sustainment cost question. I don't have a definitive estimate for the cost.
What I was giving you there was the potential cost of the operations. We don't have the sort of nailed down number for what it's gonna cost yearly. We and
Speaker 8
Well, I appreciate you saying that because eventually, you the secretary is gonna have to come back to the committee and ask for a substantial amount of money to sustain the EHR. Correct. And, I mean, we're almost 50,000,000,000 into it, it's not even working. Right. That's gonna be a very interesting conversation.
Speaker 5
Right. And there will be part of that conversation is also what money comes off the books as we don't have to sustain, you know, legacy technology that we will be able to shut down at that point in time. So, but coming back to how sites were selected. We started, we spent a lot of time looking together to say what sites have the highest level of readiness to move forward. And part of that was based on where we had already made investments.
So the sites that are the earliest sites in our schedule that we provided to you through 2031 were the sites where oftentimes we had started deployment activities and we had already made an investment, and we could save resources essentially by getting restarted there earlier. We'd already done the infrastructure upgrades at those sites. We're ready to roll. We chose to move forward, as I mentioned, with a market-based approach. One of the things that we've learned, lessons learned, is that it is better for sites in a region to all be using the same electronic health record.
That is, if you take a look in Michigan, there's a lot of interdependencies between the VAs there. You look at Ann Arbor as an example. I think it's more than 20 percent of the patients they see in Ann Arbor have a primary care provider elsewhere in Michigan, for example, in Battle Creek or Saginaw, and come there for specialty care. Having those sites on the same going. Is this?
Speaker 5
Same electronic health record allows us to have some efficiencies. And so we were looking by market, going one VSN at a time. We're gonna complete by 2026, VSN 10, and then we will go to VSN 12 and 23 and beyond from that based on the schedule.
Speaker 2
Alright. Thank you. I am going to recognize myself for five minutes, for another round of questioning. Miss Harris, I started to ask you about the kind of simultaneous testing that is going on, that is, you know, the GAO has raised as a concern. Miss Irma, what do you feel is a counterpoint to that?
And what can you do to convince me and the GAO that doing this in a simultaneous fashion is the appropriate way to do it without just hitting a deadline for a date, without thinking through the risks associated with it?
Speaker 6
Sure. Well, first of all, we strongly support, you know, robust testing. That's an important part of the implementation. When you do testing, that's when you may see things that you didn't anticipate. So agreed that it's a very important part of an overall deployment.
I think it's important to recognize in this situation that the system is live in six sites today. And since we have implemented in those six sites, we've also done a number of optimization projects. And so in though when we do those optimization projects, we go through a big process of testing the system. We do that with the VA. And then what we've also more recently implemented is that instead of waiting till the end to do testing, we kind of do it on an ongoing basis.
We show them the optimization work. So I think we've had very strong, robust testing. And so we feel confident that the system that we have in place around testing is working. There have been improvements. I think an IMV vendor at this point would just add to cost and not necessarily add anything new.
We have had now successful implementations through our optimization work that show that the testing is working. There are times when we've done testing, we've picked up things, we've delayed in some cases when we realized we needed to spend more time on it. But I think at this point, the IV V vendor is.
Speaker 2
But laying aside the IV and V, just doing all four sites simultaneously instead of one after another does increase the risk if something is discovered that needs action. Correct? And increases the risk of potential problems. Would it not?
Speaker 6
Right. This is more of a scale issue. When we know we have four sites, we make sure that there's adequate teams on the ground to do elbow testing. We also have a war room so that our teams are actively monitoring.
Speaker 6
Even before that. Even before that. Right? That's when doctor Evans and I spoke about hiring enough staff to make sure that we can scale, so that we are testing. We're supporting those sites as that's going on.
So it's just a matter of scale. The process is the same around testing, around training, and around supporting. It's just a matter of having enough staff to do it. But this is each.
Speaker 2
Of these are very unique in their application. Each of these VA hospitals, they're not, you know, cookie cutters stamped out. These are very customized to their unique situation, and doing them all four together to me elevates the risk that there's going to be problems that arise or issues that are overlooked. Or we're batching these all together, and then we're gonna have them all go live nearly simultaneously without a lot of, you know, without a lot of consideration for what happens if, as mister Latrell pointed out. And I think those are the things that we have to bear in mind.
And, you know, not to discredit what you're saying, but it feels like a lot of that is, you know, we have these things and, you know, we're a big company and we can do that. But I don't think that Oracle has had a project like the one that the VA is undertaking right now. Would you agree with that?
Speaker 6
I think we feel very confident that we can do this, and we can do it at more than one site.
Speaker 6
We feel very confident in doing that. And I agree with you that every VA is different, which is why we've had teams on the ground. It's not like we're turning on the system quickly. We've been in these sites. We've been doing assessments of the sites so we can understand what they have on the ground, what the differences are between each site.
So we have a plan for each site, and we have adequate staff and support for each of those sites.
Speaker 2
Miss Harris, do you feel like these commitments by Oracle are satisfactory to the overall concerns that GAO raise?
Speaker 7
I mean, in taking a look at the previous history of the initial six sites, particularly in the five, I mean, when they went live, Oracle Cerner did have a difficult time in addressing those ticket resolving those tickets in a timely manner. I know that they did a lot of streamlining in that process so that they would be able to meet their contractual marks for completing or resolving those tickets, you know, against their contractual obligations. But I think, again, when you're doing it for simultaneous ones, I mean, there's a tremendous amount of resource that are gonna go towards ensuring, for example, that that ticket resolution is done, you know, under the contractual obligation. So that alone, I think, is very risky. It's gonna take a tremendous amount of resources that I'm not quite sure are sustainable for multiple sites a month.
Speaker 2
Last question. Is that something that was at the request of the VA or Oracle to do these four simultaneously? If I may, either one of
Speaker 5
you. Yes. It was VA. VA asked to do this. And I would say, I'd like to take us back a step.
One of the parts that we have talked about in this committee that is super important for us to succeed is to standardize our workflows. And one of the things we've done over the last year is establish very clearly what the federal EHR baseline is. As a reminder, this system is one instance, one system to support all of these medical centers. Part of being locking down a baseline will actually streamline our ability to do testing, because some of that variation that you're mentioning from one site to the next becomes less, and testing becomes much easier at scale when you're testing against a standardized set of workflows and a standard baseline. The other, the second thing is that we are, you know, as our
Speaker 2
sorry. I gotta yield to the ranking member, but we will come back for more questions, time permitting. Ranking member Buzynski is recognized for five minutes.
Speaker 4
Thank you. I actually just wanna pick up on some of your questions about readiness. I was curious, doctor Evans, because I think you and Miss Velma have talked a lot, pointed to North Chicago as the example of how of readiness. But I am curious of how well the readiness is going there and the development since it's go live. That's a unique case, as we know, though, because that was with DOD.
Can you speak, though, to the other sites that have already gone live as well? And just, you know, what's happening at those sites as it relates to readiness?
Speaker 5
Sure. At the sites other than FHCC, which are in Spokane, Walla Walla, Roseburg, White City, and then also Columbus, Ohio, I would say overall, we are tracking numerous metrics at those sites. You heard Ms. Firma mention about the revenue capture. We're doing very well with regard to appropriate revenue capture.
Productivity at all sites at Roseburg has returned to greater than the 2019 pre-pandemic baseline productivity. We continue to hold a problem management forum with live site recommended representatives every single day, every single workday, where issues are surfaced. We address those. We've done a lot of work to work down the ticketing backlog for significant change requests, reduced that by well over 40%, and are being quite responsive to the sites with their needs. Frankly, they can escalate anything any day at our 10 AM meetings.
So we're seeing definite improved operations with the EHR at those original five sites.
Speaker 4
Can I just ask on those sites? How are you measuring the productivity and how it's changed at those sites?
Speaker 5
For the productivity, at those sites, are looking at something called RVUs, which is a mechanism. It's an industry standard for measuring productivity. It's captured through billing encounters or encounters that sort of capture the documentation or the work that was done at the time of visit. We look at our views per provider.
Speaker 4
Mhmm. Okay. I'd like to shift to Ms. Harris. So it was, we talked about the testing. I also just wanted to circle back to your testimony about the GAO recommendations, more broadly, that are still open.
What do you think are some of the most concerning of those that are still open today? I think one of
Speaker 7
The most concerning one is change management. VA still has yet to have an approved and formal policy in place there. As part of change management, there's training as well. When you take a look at the first five sites as well as FHCC, those are vendor-led trainings. And at all of those sites, the feedback has been that those vendor-led trainings had failed to prepare them for their specific roles and workflows.
And so I think it is important for VA to take a more leader role in that training. So, again, we have those open recommendations there. That's vitally important.
Speaker 4
Okay. And, doctor Evans, in addition to those recommendations that Miss Harris identified, I think there are 28 open from VA's inspector general as well. What is the plan and timeline to close them?
Speaker 5
With regard to the JAO recommendations, we take, of course, all of the JAO recommendations seriously. Of the 17 open JAO recommendations, I do wanna highlight that for eight of those, nearly half, VA has made very significant progress. And JO has asked to keep those recommendations open until after we get past some of these initial go-lives in Michigan specifically. So for example, there is an open recommendation about making sure that our contractor staff and the VA is using the right terminology. This is a recommendation from 2020.
We fully implemented that recommendation. JO asked us to wait to see how things go in Michigan, to close that recommendation.
Speaker 4
I just interrupt? Because I'm running out of time. But, Miss Harris, would you agree with that? I did note that some of the recommendations do say partial, but many of them still say open. So, is there more to the story, from what Doctor Evans is saying?
Speaker 7
There has been action taken on many of the recommendations for sure. The one that doctor Evans noted is not one that we do consider to be priority. It is important, of course, but certainly, all 12 of our priority recommendations, those are still the majority of them is not in the situation that he had mentioned. However, I will say they've done work, but we do need to hold some of those open to ensure that the actions that they've taken are effective, because we'll see those results at Michigan. Okay.
I yield back.
Speaker 8
Thank you, Mr. Chairman. The six active systems say that the system itself has proven difficult to use and is not well suited for VA workforce workflow. Excuse me. Staff have reported slow performances, excessive clicks, data loss, and a cumbersome documentation process. All six facilities that implemented the new system have faced increased workloads due to workaround processes, burnout, staffing shortages.
The report goes on about talking speaking on because of the weight of the implementation of this, morale is down, and there have been system updates. There has been a standardization and readiness improvement. And the user experience apparently went from 7% in '22 to 33% in '25. Now these are the six sites that are currently working under the Oracle system. And you're about to add 4 more.
Mister Evans, you said in your previous statement that all the all you're doing Michigan because Michigan's the same. We're gonna do Washington because Washington's the same. You come down to Texas, Texas is Texas. Now if you have these particular sites that are having these issues, you're gonna add these other sites on board that are different, and then you're gonna try to complete the entire system. That's of extremely heavy weight.
I'm curious, is Verma or doctor Evans? What does that even look like? Because if the current system doesn't work effectively at 33%, and we're about to add four this next year or four in April, and then what's the number in 27, please? 26 sites. 26 and 07 or 27. You see where I'm going with this?
Mhmm. Walk me through it.
Speaker 5
I do. First of all, I think we have just to give you a sense for the change over the course of the reset. We introduced over 1,500 functional changes. That is changes responsive to end users where end users said, hey. The system needs to work better in the following way.
Speaker 5
1,500 changes over the course of the reset. That was more than 50 changes on average a week that we were introducing, responsive to their feedback. That's not beyond that, thousands of other changes that were introduced as part of block upgrades, which are platform upgrades where Oracle's improvements to their base platform are introduced to the VA. There has been a significant amount of improvement to the user experience. And I think the data that you are quoting, the 33%, is from our end-user experience survey.
And, I believe that we are seeing a direct line improvement. That is as we pay attention to end users, as we address the change requests in a standardized way at the national level that is responsive to our clinical communities, and as we deliver those in a way that, as Miss Harris mentioned, is rolled out in a way where users feel supported, where they get adequate communication, where they get training to the changes. We are seeing confidence improve. Now it's not where I want it to be, but we are seeing confidence improve at the six live.
Speaker 8
sites. Are the four active sites in Michigan? Is this implementation has this been handed off to them so you won't be surprised when something like this populates after they go active 117 days. They're not gonna say hopefully, no I shouldn't. I'm not gonna call it a mistake, but they're not gonna have the same issues that the current sites are having because everything that we've learned off the current six sites has been pushed over to the four sites that are gonna go active in April.
Speaker 5
That is correct. All of the improvements that the six sites that we have learned from, the six sites, and frankly, from our lessons learned working with the DoD and working with Oracle and commercial customers about what best practices are, all of that value will be delivered to Michigan. Hopefully. Yes. It will.
I yield, sir.
Speaker 2
Thank you, Mr. Latrell. Recognize myself for another five minutes. So I know this issue of change management and everything is difficult. And, doctor Evans, I think I shared with you that I think, as a classification, physicians tend to be a little bit stubborn, and I think you agreed with me. And then I think physicians within the VA might be a particularly stubborn bunch.
But with that being the case and laying that aside, who's ultimately responsible for the change management? Is it Oracle? Is it VA? Or is it Accenture? Like, who is responsible for that part of this?
Speaker 5
I mean, at the end of the day, VA is responsible. This is our health care system. This is the VA health care system, and this is our project to implement a new electronic health record. One of the things that we, as we during, you know, as we sort of sat before we started to accelerate deployments, we said we need a new change management strategy. We worked that out.
One of the big pieces of that is that it is VA leaders standing in front of their peers that kick off this process. We introduced a new event. It's the change leadership team, executive leadership team onboarding event. We do it in every market. It is led by VA leaders who are speaking as peers to their staff to say, this is what the change will feel like.
This is what it's going to be. I can own a little bit of stubbornness as a primary care provider in the VA, as you pointed out. But I will say this, my experience is that VA clinicians circle around the veteran. If you are delivering a better experience to the veteran, that is a worthwhile change to adopt. And part of this is building a real sense of the why.
Why is there value in us coming together around a single electronic health record that supports veterans wherever they are, where the care surrounds the veterans, regardless of what physical geographical location they walk into in a VA? And I'm not seeing objections from our clinical staff or our administrative staff when they understand why and when they're adequately supported. So we own it. VA owns it. But I am very grateful as well for the contractors who are supporting us in executing that change management task.
And.
Speaker 2
Then, the surveys that you're undertaking, you know, mister LeChow point out you guys went from 7% to 33%. So, congratulations. You're more popular than Congress, finally. But with that being the case, do you feel the survey methodology is accurate, or is it one of those things that tends to attract more negative response bias?
Speaker 5
Yeah. That's a good question. We get about 20% participation in the surveys on average when we send them out twice a year. And that's actually a good response rate for a survey when we're asking busy folks in the medical centers to take their time out to do the survey. Point number two, the survey methodology: the survey questions that we use are a standardized set of questions used by many health care systems, both in the private sector as well as in the federal government, and are comparable with the DOD.
And so those questions, I do believe that they are robust questions that we can learn from and by and from which we can look at other health systems that have engaged in a similar transition and track our progress accordingly.
Speaker 2
Okay. And do you feel that VA end users, as they're being trained on this, have adequate, you know, authority to raise their concerns without feeling like they're being, you know, that there's gonna be not punishment, but just, you
Speaker 5
I mean, it is an anonymous survey. We expect, yeah. Complete honesty on the survey.
Speaker 2
Okay. Switching briefly because I've only got about 1 minute left. I know we talked, Miss Firma, a little bit about some of the pharmacy-related implementation and things like that being a unique role that the VA has and some of the drug interaction pieces. What, I guess, where do you see Oracle integrating into this to make sure that we can resolve that going forward for the understandably unique way that the VA does the pharmacy role?
Speaker 6
Sure. And you're absolutely right. Right? We're taking an off-the-shelf solution and bringing it to the VA, and the VA has some very specific and unique needs. I think that there's been some points that we've implemented or some projects that we've implemented that have really upscaled the level of safety, things like opioid prescribing.
We've also improved communication between pharmacies and the providers if a drug's not available, to have that conversation so that we make sure that what's being prescribed is available. And we are seeing some definite improvements. We are seeing some providers out there that have a level of productivity to pre-deployment. That being said, I think we would recognize that this is an area of continued focus where we want to make sure that we have adequate training. We also sent, you know, our CEO went to go visit one of the centers and specifically looked at pharmacy because we do know that this is going to be something that we are going to continue to improve.
So I think we've made progress, but I would also acknowledge that this is an area that we continue to focus on at the highest levels of Oracle, and we continue to make improvements. And we have conversations about this on an ongoing basis. One of the things that I really appreciate is the leadership team.
Speaker 2
Cut you off in just a second. I apologize. I'm recommending for 5 minutes.
Speaker 4
Thank you. Miss Firma, Oracle has been touting its effort to build a new EHR. I think you've gone as far as to say that Cerner's EHR is equivalent to crumbling infrastructure. Is this crumbling infrastructure the product that is being deployed at the VA?
Speaker 6
Improvements to the Cerner system that's being implemented in the VA. So we took the Cerner system. You've heard about all the different optimization projects that we are doing. We are also introducing a lot of AI agents that will sit on top of the Cerner system to help it provide, you know, the best experience for providers on the front lines.
Speaker 6
I think a lot of it is informed by lessons learned. One of the things that we appreciate is that the leadership today is very active in terms of visiting the sites, seeking input from providers, understanding what the pain points are, and then communicating it to us. We meet very frequently with the VA, almost every two weeks. There are weekly meetings with the secretary. And in those meetings, we're identifying where there are opportunities for optimization.
We've executed a lot of these projects. You've heard of the capability block updates that we've been doing, and those represent the optimization projects. And I think that's why we are starting to see better results in terms of the experience of providers. And I think the sentiment overall is increasing as well, in terms of the positivity.
Speaker 4
Of time today talking about cost, and I'm just curious how you'd respond. You know, hearing this, is this going to all lead us down the path of additional cost from Oracle's perspective?
Speaker 6
Yeah. So I haven't reviewed any of the new estimates that have come out, but I think that's something that we're very cognizant of with cost. I think that some of the more recent changes that we've made, the first one I would say, is the standardization. You know, having a system where you had six different sites requesting different types of changes, those kinds of things contribute to increased costs. So we're very excited about the changes that we've heard about in terms of moving to standardization.
That's going to make the cost more predictable and more sustainable. That being said, it's always, you know, foremost on our minds. This is why Oracle agreed to move the federal system to our cloud to make sure that we had, you know, greater security and performance, and we're doing that at our cost as well. And we advised the VA. One of the things that we really advocated for was the standardization because we also knew that not being standard would contribute to increased costs.
I think the pause in and of itself hasn't helped in terms of cost, because you're maintaining two systems. And so moving forward, I think, will also help make sure that we are using taxpayer dollars appropriately and efficiently.
Speaker 4
And is, we've talked, you've talked a little bit about in your testimony, obviously, AI. Is AI capability included in VA's contract with Oracle? And is there an additional cost to that?
Speaker 6
It depends on which AI agent. So there's not a one-size-fits-all approach. We are bringing the brand-new EHR to the VA. We are not charging for that, you know, the new ambulatory system. But there will be some AI agents that are included inside the EHR.
There are some that are additional. And the BA will have to assess those and decide which ones they want to use.
Speaker 4
Okay. Ms. Harris, can I just ask you, in your testimony you'd flagged, we were talking about the overall cost at the beginning of this hearing? And I didn't want to get to you. But could you, I think you had flagged some concerns around that number, the $37,000,000,000. Could you speak to that?
Speaker 7
Yes. Well, we haven't seen the 37000000000 number. So as soon as our office receives that, we will certainly do a deep scrub of that, and then get back to you on that. But I will say the independent cost estimate that is out there is the total life cycle cost is roughly $49,800,000,000. So, Mhmm.
Roughly $50,000,000,000. And so, you know, we're gonna have to go through the differences. That number also is outdated because it doesn't reflect the changes in the delays, including the pause. So that that's also something that, we'll have to take a look at, but we'll certainly do a, a comparison there of that. Okay.
Speaker 4
That. Thank you, when you get that. Miss Verma, if I could just ask again, going back to Oracle Health software and specifically the veterans' data, is it being used to train the AI that Oracle will use for its commercial clients?
Speaker 6
Absolutely not. We do not have access to any data rights, so we do not use that data to train our models. No.
Speaker 4
Okay. Will VA receive credits of some sort for the benefit or goals derived from the information?
Speaker 4
Okay. Okay. Miss Harris, your testimony indicated the VA has not instituted plans to conduct an independent operations assessment to evaluate the suitability and effectiveness of Oracle's EHR. We've been contemplating the need for such an evaluation. What do you think this should entail, and who do you think should conduct it?
Speaker 7
Yeah. So I think that's, given this approach to go live at four sites at Michigan, once that takes place, I think that there should be an IVMV after that looking at all four sites to do that systematic cataloging of those defects. And it should be an independent third party that goes in and does that review. And, again, I can't stress enough the reason why MHS Genesis was so successful in their deployments. You know, we've spent a lot of time with them.
A large part of that was because of the IV and V tests that they performed.
Speaker 2
Sorry. Thank you very much, Miss Slotrel, for five minutes. Then we'll look.
Speaker 8
Thank you, Mr. Chairman. So the VA did not doctor Evans, don't know how long you've been in the VA, but I was curious. So we didn't negotiate in our contract with Oracle any kind of AI footprint. Because what I heard her say is we're just gonna get to baseline. It's like this is like when my eight-year-old comes up and wants to buy a video game.
He's like, hey. We can get the baseline model. Then when you're inside, you gotta buy everything else to get it to where it needs to be. It kinda sounds like where we're at. You don't have to answer that question.
I just wanted to say that out loud. We have six sites that are currently on Oracle. Right, Miss Verma? So the rest, 160 plus, are still on what? Vista à?
Doctor Evans, we spend roughly about 1 billion dollars on software across the board. Correct? Some of the VA facilities use software that other VA facilities don't use. Some have the highest level of software, advancements inside their facilities, and some don't. But we have to pay for it all. We're not even using it or not.
Like, my little facility in Conroe, Texas still uses some software that helps Vista a, but DeBakey does not. But we still have to fund the whole thing, and it's almost 1 billion dollars. I'm correct. I may be off on that, I think it's almost a billion dollars. This vermin, when Oracle activates on every single site, do we have the ability to tell everybody else, hey.
We don't need you anymore.
Speaker 8
It can depend on every site because the whole purpose of us doing this with you guys is that every site is the same. Yes. So my says that multiple times.
Speaker 6
Yeah. So my expectation would be that with a new Oracle EHR, they should be able to use our system and that should address the needs that they have. I don't know whether, you know, every single site and what they have, but I would anticipate that this EHR with the standardization that we're bringing, that it should be able to meet the needs of any hospital or clinic across the country.
Speaker 8
I hope so for $50,000,000,000, because we're gonna have to have the conversation with everybody saying, hey. Look. We implemented this at $50,000,000,000, plus the dollar sign, Doctor Evans, of what sustainment looks like. And then, hopefully, at the end of the day, we don't have to say we have to continue to pay for everything that's in Vista because we already have it in the system, and then we're gonna pay for Oracle on top of it. Are we gonna have to have that conversation?
Miss Harris, do you think about that? That sounds reasonable? This is kind of the math problem I'm drawing out in my head, but it's where we are. The slide deck that they brought to my office shows every single software that we pay for. We have one software program that only one VA uses, and we still pay for it because they have to have it for some reason. What do you think about that, doctor Evans?
Speaker 5
I think that, you're, this is a very important point. As we move forward, the federal EHR, think of that as the operating system for the hospital, will be the same at all hospitals. Not all hospitals offer the same clinical services. For example, we're going live in Michigan. For the first time, we'll be implementing radiation oncology.
There are some unique software that is necessary to support a radiation oncology clinic and operation that won't be needed at every site. But one of the real advantages and one of the reasons why when you hear me and us and VA talk about what we call the federal EHR baseline, we think that is so important, is because we are going, we are publishing. It's published right now on our website. These are the software systems that we will support and can support connected to the federal EHR. And by definition, it also means there are those that we won't support.
Speaker 8
So every system that we have in place currently, once Oracle activates in all sites, is gonna come running to you guys and say you have to use this?
Speaker 5
Well, again, because we're already here? No. No. The we are making the decisions now as we go from site to site around what the standard is going to be.
Speaker 8
I'm sorry. I'm still on the impression that every site's gonna be the same.
Speaker 5
Not every site has the same clinical services. I understand that. But when we go to a site.
Speaker 8
But every site can have what Oracle is populating in the six sites that are active, and four sites that are coming up, and the 27 next year will be the exact same. Correct? So I have that in my little satellite campus in Conroe.
Speaker 5
So I'll give you an example. Bedside monitors, when you walk into an intensive care unit, the thing that's hanging up above the bed with the EKG on it, that is not a core part of the electronic health record. That is a biomedical device. It has software that runs it that we'll need to plug into the electronic health record. One of the ways we will be able to manage cost is by standardizing which of those devices we can support.
And it should be a limited number in order to be able to manage the EHR in a more cost-effective way across the enterprise.
Speaker 2
Thank you. I apologize. We're up against a bit of a hard stop for the committee. We're gonna move on to closing statements. Do I go to you first?
Okay. I'll go to ranking member Buzinski for her closing statement.
Speaker 4
Thank you. I agree that we need to give veterans and VA employees the modern tools that a new EHR can offer, but those tools have to work for the VA. And I just don't believe that we're there yet. I don't wanna be a pessimist, but I don't feel like I'm leaving this hearing having my mind changed on this point. Doctor Evans and Miss Firma are quick to point out that the new administration is driving increased momentum and leadership involvement in the project.
I'm concerned that the administration's involvement is only moving the program forward faster, not better. There are so many recommendations from the GAO and the inspector general, as well as Congress, that will continue to sit unheeded. I have no confidence that the next round of go-lives is gonna be any better than the last. I would implore Secretary Collins and the Trump administration to pay attention to their own words and put the veterans at the center of everything. Anything else puts the health and safety of our veterans at risk.
Thank you, and I yield back, Mr. Chairman.
Speaker 2
Thank you, Ranka member Buzinski. I was actually voted most optimistic of my high school graduating class, mostly because I thought it was the Lion's Year every year. And, again, we find ourselves on the edge of not even making the playoffs, perhaps. But I, in my old age, I've moved from an optimist to more of a realist. And I think the issue before our committee is what is the real assessment of where we're at, and what we need to do to prepare going forward.
And I think a healthy amount of question, concern, and even a little bit of skepticism is appropriate for our committee to feel, given the past performance of how this has gone and what we need to be prepared for. And I can tell you with candor, what I will not accept if we fast forward the tape a hundred and seventeen days from today is if things don't go well and do not go as necessary to protect the health and benefit and welfare of the veterans that we serve is finger pointing and blame between, you know, various different vendors, the department, and who had what and musical chairs as to where things land. So this committee, our work here is not gonna tolerate that. If there are differences between opinions as to what needs to take place, I encourage you to resolve them. If you need help from this committee to do that, I am very willing to be the person to step into that role and with the ranking member to assist in doing that.
But we need to absolutely make sure that we have a no-fail mindset going into this go live, 117 days from today. When I was at my last appointment in Battle Creek for a physical, they set a follow-up that was a few months into the future, and it happened to land on the week that this go live is. And, you know, I'm comparing my session calendar with when I can be back home, and I'm like, well, I'm available on this day. They're like, oh no. We are getting a new electronic health record system.
We are not scheduling appointments on that day. And so I know that they are taking it seriously. I just wanna make sure that the tools are gonna be there for the end user, the practitioners, the doctors, medical assistants, the nurses, and all the other various people involved in this from front to back are going to be ready for this as it comes up. I'm hopeful and encouraged that you pointed out the superusers have begun getting their training now, but then the more, you know, rudimentary kinda day to day users are gonna be getting that at an appointment that is appropriate for them. I have concerns about the, you know, simultaneous testing going on as we pointed out in the questioning, and I wanna make sure that is done appropriately.
And I think there are questions that the committee has that are still unresolved, and a few more that Mister Latrell raised as well. So, we wanna be partners in this, not antagonists, and we don't want to slow you down. But we do want to maintain the appropriate level of accountability to make sure that this goes as well as we need it to. So, with that, I think I have some disclaimers I gotta say here at the end. I ask unanimous consent that all members have five legislative days to revise and extend their remarks and include extraneous material. Without objections, so ordered.
And with that, this hearing is adjourned.
Ask anything about this video — I analyse its transcript to answer.