Summary

  • MEDITECH held 14.7 percent of US acute-care EHR hospitals at end-2025, down from 14.8 percent, with a net loss of seven hospitals and 1,015 beds during 2025, per KLAS data reported by Fierce Healthcare; it remains the third-largest acute-care customer base behind Epic (43.7%) and Oracle Health (21.9%) (Fierce Healthcare).
  • Retention among legacy customers reached a record: 84 percent of the 44 legacy MEDITECH customers making a go-forward decision in 2025 chose to migrate to Expanse, up from 63 percent in 2024 and 30 percent in 2023 (KLAS 2026 blog; HIT Consultant).
  • The market itself contracted: hospitals impacted by an EHR purchase decision fell roughly 40 percent versus 2024, and only Epic and TruBridge gained net share in 2025 (KLAS Research).
  • MEDITECH's flagship offensive win is HCA Healthcare, where Expanse is live at 43-plus hospitals with full US divisions live, in a rollout that industry commentary in September 2026 referenced at around 72 hospitals (MEDITECH; Becker's Hospital Review; healthsystemcio).
  • KLAS analysts report that MEDITECH customers and prospects still cite a need for better interoperability and broader native capabilities to eliminate third-party add-ons — the same gap that lets Epic grow among small systems (Fierce Healthcare).
  • Over 400 hospitals remained on MEDITECH legacy platforms at end-2024, about 45 percent of its acute-care base; this backlog is simultaneously MEDITECH's greatest retention asset and the ceiling on its growth (HIT Consultant; Fierce Healthcare).

The numbers behind the paradox

The 2025 market-share figures, drawn from KLAS's 2026 report and summarized by Fierce Healthcare, put the US acute-care EHR market at 43.7 percent Epic (up from 42.3 percent), 21.9 percent Oracle Health (down from 22.9 percent), 14.7 percent MEDITECH, 7.6 percent TruBridge and 2.9 percent Altera Digital Health, with the remainder distributed among smaller vendors (Fierce Healthcare). Epic gained 77 multispecialty hospitals and 18,679 beds in 2025 and lost a single customer. Oracle Health lost a net 56 hospitals and 14,676 beds — its third consecutive year as the market's largest net loser — with many customers deferring decisions while they wait for clarity on the vendor's direction. MEDITECH, by contrast, lost seven hospitals net.

That asymmetry matters. MEDITECH is not bleeding to the same diagnosis as Oracle Health. Its 2025 net loss is roughly an order of magnitude smaller than 2024's, when the company lost a net 57 hospitals — mostly because one large health system moved 41 hospitals from Expanse to Epic. Of the 23 organizations that competitively replaced MEDITECH in 2024, 19 chose Epic, frequently citing interoperability benefits with nearby Epic organizations (HIT Consultant).

Meanwhile the conversion engine inside the base accelerated. In 2023, only 30 percent of legacy customers making a go-forward decision chose Expanse. In 2024 that doubled to 63 percent. In 2025, it reached 84 percent of 44 deciding customers — MEDITECH's strongest retention year yet, per KLAS (KLAS 2026 blog; HIT Consultant). KLAS attributes the improvement to strong client partnerships and to MEDITECH as a Service (MaaS), the vendor's subscription hosting model (KLAS 2026 blog).

The catch is arithmetic. About 45 percent of MEDITECH's acute-care hospitals still run a legacy platform, and more than 400 hospitals were on older MEDITECH systems at end-2024 (Fierce Healthcare; HIT Consultant). Every legacy conversion is a retention win but not a market win: a converted hospital was already MEDITECH's. The 2025 result — record conversion, negative net share — is exactly what that arithmetic predicts.

What Expanse actually automates

Expanse is MEDITECH's web-based, mobile-capable successor to its client-server 6.x platforms, spanning acute inpatient, ambulatory clinics, emergency, anesthesia, oncology, behavioral health and home care from a single database. The vendor's recent product push centers on Expanse AI: ambient documentation that drafts clinical notes from patient-provider conversations, retrieval-augmented search across the record, and AI-assisted coding and summarization features that MEDITECH says are designed to run inside the existing Expanse license rather than as separately sold point products (MEDITECH, Expanse AI).

Two properties of this automation strategy distinguish MEDITECH from its larger competitors. First, breadth-within-one-system: the pitch to rural and community hospitals is that charting, ordering, scheduling, revenue cycle and now ambient documentation come from one vendor on one database, reducing the integration surface a small IT staff must maintain. Second, deployment economics: under MaaS, the customer buys the software as a subscription with hosting, support and upgrades bundled, rather than funding a capital license plus an on-premises hardware lifecycle. MEDITECH's own materials and partner analyses position MaaS as the counter-cyclical option for capital-constrained community hospitals; the customer trades long-run licensing flexibility for predictable operating expense and a vendor-run upgrade cadence (MEDITECH, MaaS).

Expanse also runs on Google Cloud, where MEDITECH is a named customer case — a hosting posture that reduces the vendor's dependence on customer-owned server rooms and, functionally, makes MaaS deployments indistinguishable from cloud deployments to the end user (Google Cloud). For a 50-bed critical-access hospital with two IT generalists, that trade is frequently decisive: the alternative is Epic, whose community offerings are priced and staffed for larger organizations, or a smaller vendor whose future is less certain.

The HCA exception proves the constraint

If record legacy retention were the whole story, MEDITECH would be a defensive vendor with no offensive capability. The HCA Healthcare agreement is the standing counterexample. HCA — the largest US for-profit hospital operator — signed a new agreement for a large-scale Expanse implementation, and by 2026 Expanse was live at 43-plus HCA hospitals with MEDITECH announcing the first full US divisions live (MEDITECH). Becker's reported 32 hospitals live at an earlier milestone, and industry commentary from September 2026 discussed the rollout at roughly 72 hospitals (Becker's Hospital Review; healthsystemcio).

The strategic significance is not the headline count; it is what the deal demonstrates about MEDITECH's ceiling. HCA chose Expanse in divisions where the alternative was standing systems, and the rollout shows MEDITECH can execute multi-hospital, division-scale implementations on schedule — a capability KLAS's large-system segment usually reserves for Epic and Oracle Health. But it is the exception that illuminates the rule: in 2025, only two US health systems with more than ten hospitals made enterprise-wide EHR purchase decisions, and both chose Epic, primarily taking hospitals out of Oracle Health (Fierce Healthcare). Large-system displacement of Epic by MEDITECH has no comparable example in the KLAS record. MEDITECH's offensive market is the independent community hospital and the mid-size system, where its product scope, MaaS economics and implementation cost structure are matched to the buyer.

The interoperability gap as competitive fact

The clearest statement of MEDITECH's constraint comes from KLAS's own analysts: MEDITECH customers and prospects cite a need for better interoperability and broader capabilities to eliminate third-party solutions, for example in capacity management (Fierce Healthcare). This is the mirror image of Epic's growth mechanics. Epic's dominance among small systems in 2025 was driven partly by interoperability gravity — hospitals want to be on the same platform as the referral networks around them. Of the 23 organizations that replaced MEDITECH competitively in 2024, 19 chose Epic for precisely this kind of reason (HIT Consultant).

MEDITECH's responses are visible in its product line: Expanse AI features, a persistent vendor-managed hosting path, and 12 consecutive years of Best in KLAS recognition for overall EHR satisfaction, which the vendor cites as evidence that customers are not leaving out of dissatisfaction (MEDITECH, Best in KLAS). The 84 percent conversion rate corroborates that reading — legacy customers are not abandoning the vendor, they are consolidating onto its current platform. The unresolved question is whether a satisfied base plus AI differentiation can reverse the net-share arithmetic in a market where purchase decisions fell 40 percent in a single year (KLAS Research).

Sources