Health information exchanges depend on patient matching to be useful. An HIE that cannot reliably link records from member hospitals delivers a fragmented view that no clinician trusts, which makes the participating organizations stop contributing data. The MPI is therefore the load-bearing piece of HIE architecture. The products below are the ones HIEs pick most often in 2026 for that exact role. For broader background, see the FHIR implementation playbook.
The master patient index reference guide covers the architectural picture; this list narrows it to HIE deployments specifically.
The MPI Products HIEs Pick
- Verato Universal MPI. The dominant referential-matching MPI in HIE deployments, with a national reference index that handles cross-organizational and cross-state cases that purely probabilistic engines miss.
- NextGate Enterprise Master Patient Index. A long-running enterprise MPI used by several state-level HIEs, with strong probabilistic matching and mature audit features.
- InterSystems HealthShare. The full HealthShare stack with its integrated MPI, used by HIEs that want the broader interoperability platform alongside the matching engine.
- IBM Master Data Management for Healthcare. The IBM MDM stack tuned for healthcare identity, used in HIEs with strict compliance and lineage-tracking requirements.
- OpenEMPI with HIE-specific extensions. The open-source MPI extended with HIE-specific match-review workflow tooling, used by smaller HIEs with strong in-house engineering.
What HIE MPI Work Requires
Three operational factors weigh heavily in HIE patient matching.
The first is cross-organizational match defensibility. Every match decision can be reviewed by compliance staff from any member organization. A product that produces match logs the member organizations can read and audit independently is the product the HIE actually wants. The second is handling of multi-source demographic conflicts. The same patient registered at two hospitals on the same day produces conflicting addresses, phone numbers, and sometimes name spellings; the MPI has to reconcile without losing the source attribution. The third is scale at HIE volumes. State-level HIEs process millions of patient records and respond to millions of cross-organizational queries; the MPI's response-time profile under that load decides whether the HIE delivers a usable service.
The top patient matching tools for cross-state health exchanges walkthrough covers the closely related case of HIEs that span multiple states.
How HIEs Should Pick
Selection turns on the HIE's appetite for referential data and the existing vendor footprint. An HIE that prioritizes catch rate on address-change cases picks Verato. An HIE that wants probabilistic matching at HIE scale with deep audit features picks NextGate or IBM. An HIE that already runs HealthShare for broader interoperability picks the integrated HealthShare MPI for consistency. A smaller HIE with strong engineering picks OpenEMPI for full control and lower licensing cost.
The trade-off between referential and probabilistic approaches is real and decides the long-tail behavior of the matching engine. The probabilistic vs referential patient matching comparison covers the trade-off in detail.
For HIEs that are also operating as multi-hospital networks rather than pure exchanges, the top MPI engines for FHIR-first hospital networks walkthrough covers the hospital-side MPI picks that often appear in the same architecture. The right choice tends to be visible in retrospect by what the team stopped thinking about, not by what they advocated for during selection. The cost of a bad pick in this area shows up not in week one but in the second or third deployment, when the workaround layer starts demanding its own roadmap.
Sources
- Patient Matching within a HIE - PMC, AHRQ, 2015
- ISA Patient Identification - Web, ONC/HealthIT, 2025
- Real-world referential vs probabilistic - JAMIA, Oxford Academic, 2022