Ready, Set, Go-Live?

Assessing VA’s EHR Modernization Deployment Readiness

House Veterans Affairs Subcommittee on Technology Modernization Hearing
December 15, 2025

“Technology should be a tool that opens doors, not a barrier that adds more steps, more clicks, and more frustration.”
Chairman Tom Barrett

That single sentence may be the most accurate mission statement the Department of Veterans Affairs has ever had for its Electronic Health Record Modernization program. Unfortunately, it also captures exactly why Congress remains uneasy as the VA races toward another major go-live.

With roughly 117 days until four Michigan VA sites are scheduled to flip the switch, the House Veterans Affairs Technology Modernization Subcommittee convened to ask a deceptively simple question.

Is the VA actually ready?

After years of delays, a deployment pause, ballooning costs, and unresolved safety concerns, the answer offered during this hearing was cautious progress layered with unresolved risk. In short, momentum exists. Confidence exists. Proof remains incomplete.

Why This Hearing Mattered

The VA’s Electronic Health Record Modernization, or EHRM, is one of the largest and most complex IT transformations ever attempted by a federal agency. The goal is straightforward on paper: replace the aging VistA system with a modern, interoperable federal EHR that follows veterans across facilities, VISNs, and even between the VA and the Department of Defense.

In reality, the program has been anything but simple.

Since 2017, the VA has deployed the new system to only six medical centers at a reported cost of $12.7 billion. Early rollouts were plagued by performance issues, usability complaints, prescription errors, faulty alerts, and significant clinician burnout. Those problems forced the VA to pause deployments in April 2023.

Now, with that pause lifted, the VA plans to restart aggressively. Four Michigan sites are scheduled to go live simultaneously in April 2026, followed by thirteen deployments in 2026 and an accelerated path to enterprise completion by 2031.

Congress is no longer debating whether modernization is needed. The debate is whether the VA is moving faster than its evidence supports.

The Witnesses and What They Claimed

Three witnesses anchored the hearing.

Dr. Neil Evans, Acting Program Executive Director of the VA’s EHRM Integration Office, testified that the program has made substantial progress since the pause. He emphasized improved system stability, standardized workflows, stronger leadership involvement, and a renewed focus on VA-led change management. According to Dr. Evans, the system has exceeded a 95 percent incident-free threshold for 21 consecutive months, and user-experience survey scores have steadily improved.

Ms. Seema Verma, Executive Vice President at Oracle Health and Oracle Life Sciences, expressed confidence that Oracle and the VA are aligned and prepared for an accelerated deployment schedule. She pointed to reduced outages, improved productivity, expanded training, and future enhancements including AI-assisted tools, voice-first documentation, and improved interoperability through Oracle’s QHIN designation.

Ms. Carol Harris, Director of IT and Cybersecurity at the Government Accountability Office, provided the reality check. GAO has issued 18 recommendations related to EHRM, 12 of them designated as priority. As of this hearing, 16 remain unimplemented. Key gaps include updated lifecycle cost estimates, an integrated master schedule, and independent operational testing.

In other words, the VA says the system feels better. GAO says the documentation does not yet prove it is safer.

The Cost Question That Would Not Go Away

If there was one moment of bipartisan agreement, it was discomfort with the program’s price tag.

Originally estimated at roughly $10 billion, the VA now pegs the lifecycle cost closer to $37 billion. GAO has not yet reviewed that figure, and an earlier independent estimate suggested costs could approach $49.8 billion. Even VA leadership acknowledged uncertainty around long-term sustainment costs once deployment is complete.

When asked what the annual cost would look like after 2031, Dr. Evans could not provide a definitive number. He referenced a final-year operational estimate of about $2.1 billion but emphasized that future savings from retiring legacy systems had not yet been fully calculated.

For Congress, this lack of clarity matters. Oversight is difficult when nobody can confidently state what the finish line costs, or what “done” actually means.

The Michigan Gamble: Four Sites, One Switch

The most contentious issue of the hearing was the VA’s plan to deploy the EHR to four Michigan sites simultaneously.

From the VA’s perspective, this market-based strategy makes sense. Facilities within a region are interconnected. Standardized workflows should reduce variation. Shared go-live dates avoid running dual systems within the same market.

From GAO’s perspective, it is a high-risk bet with limited margin for error.

Ms. Harris warned that simultaneous deployments complicate independent verification and validation, stretch support resources, and increase the chance that a late-discovered flaw cascades across multiple facilities. Early EHR deployments struggled with ticket resolution even at a single site. Scaling that challenge fourfold raises legitimate concerns.

Oracle argued that the risk is manageable through staffing, elbow-to-elbow support, and war rooms. GAO countered that staffing intensity does not replace independent operational assessment.

Chairman Barrett summed it up bluntly. These sites are not cookie-cutter facilities, and discovering a serious problem at go-live leaves little room to adjust when four locations are already live.

Staffing, Burnout, and the Human Factor

Technology does not fail in isolation. People absorb the impact.

Ranking Member Nikki Budzinski raised concerns about reports that the VA planned to eliminate 35,000 physician positions. Even if unrelated to EHRM, staffing reductions during a major IT transition raise red flags. Dr. Evans insisted EHRM hiring was ongoing, with more than 500 positions in recruitment for Michigan sites alone, supplemented by contractors from Oracle, Booz Allen, and Accenture.

Still, members cited persistent user complaints from existing live sites. Slow performance. Excessive clicks. Documentation burden. Workarounds. Burnout.

User-experience survey scores have improved from roughly 7 percent positive in 2022 to about 33 percent in 2025. That is progress, but it also means two-thirds of users remain unconvinced.

Scaling a system that works well for a minority of users is not a comforting metric.

Change Management: Who Owns the Mess

One of the more important admissions of the hearing came when Dr. Evans stated plainly that change management belongs to the VA.

Not Oracle. Not Accenture. Not a vendor.

The VA.

That matters because GAO has consistently warned that vendor-led training does not adequately prepare clinicians for real-world workflows. The VA described a revamped strategy including market-level Change Leadership Teams, peer-to-peer executive engagement, super-user training, learning labs, and earlier hands-on exposure.

These are the right ideas. The unresolved question is whether they are enough, and whether they have been tested under real go-live stress.

Accountability Without Finger-Pointing

Chairman Barrett closed the hearing with a tone shift that deserves attention. He described moving from optimism to realism. Progress is acknowledged. Skepticism remains warranted.

Most notably, he made it clear that post-go-live finger-pointing between the VA and its vendors will not be accepted. If problems emerge in Michigan, Congress expects accountability, not explanations.

That expectation should sound familiar to anyone who has ever tried to schedule a VA appointment, refill a prescription, or untangle a medical record error. Veterans do not care which system failed. They care that care was delayed.

The Bottom Line

The VA’s EHR modernization effort is not standing still. Stability appears improved. Leadership engagement is stronger. Lessons from early failures are being applied.

But readiness is not a feeling. It is evidence.

Until GAO’s priority recommendations are closed, independent operational assessments are completed, and lifecycle costs are fully transparent, Congress is right to question whether the push is being driven by deadlines rather than data.

Michigan will not just be a go-live. It will be a proof point.

If it works, confidence grows.
If it falters, skepticism hardens.
Either way, veterans should not be the beta testers.

Technology should open doors. The VA now has 117 days to prove it will.

(quotes provided by www.thenimitzgroup.com by the Nimitz Report)