In March 2025 NextGen Healthcare announced the end of the dual-license model for Mirth Connect. From version 4.6 the engine is distributed only as a commercial product. Within weeks, community forks appeared, among them Open Integration Engine (announced in April 2025) and BridgeLink.
For an interface team running a free 4.x install, nothing broke that day. Channels kept receiving, transforming and sending. What changed is the future of the build you run: it will not get further open-source releases. That turns a quiet piece of infrastructure into a decision, and the decision deserves more than a reaction to a license quote.
This post lays out the three paths and the questions that decide between them.
Path 1: buy the commercial license
You license a supported Mirth Connect release and upgrade to it.
What you get: a supported engine with vendor releases, security fixes and a support contract, and continuity with the product your channels were built on.
Where the work is: the upgrade itself. Depending on where you start, that can mean a newer Java runtime, a database schema migration, changes in default behaviour, and re-testing every channel. Estates that skipped several releases tend to carry custom JavaScript that leans on older behaviour.
When it is the right answer: large estates with heavy custom code, organisations with little appetite for change in a clinical integration layer, and teams where the cost of a migration project plainly exceeds years of license fees. It is the right answer more often than people expect, and nobody should talk you out of it on principle.
Path 2: hold on the version you run
You stay on your current 4.x build and keep running it.
What you get: no project, no license fee and no change, for now.
Where the work is: in everything around the engine. Security patching becomes your responsibility, Java and operating-system support will age underneath you, and the knowledge of how the estate works often lives with one person. Holding is a decision to accept those risks, ideally for a defined period.
When it is the right answer: small, stable estates with a known end date, such as an engine due to be replaced by an EHR's own integration tooling, or an organisation that needs a year to fund a project properly. Holding without a plan is the one path we would argue against.
Path 3: migrate to an open fork
You move channels to Open Integration Engine or BridgeLink.
What you get: an open-source engine with a channel model your team already knows, and no commercial license fee for the engine itself.
Where the work is: compatibility and proof. Every extension, connector, plugin and script your channels depend on has to be checked against the target. Channel exports may import cleanly and still behave differently. The migration needs parity testing, a cutover plan and a rollback path, and you are taking on the fork project's release cadence and community support model.
When it is the right answer: estates that are well understood, or can be made so, where license cost matters and the team is comfortable with community-supported software.
The questions that decide it
Rather than starting with a preference, start with the estate:
- How many channels do you run, per environment, and how many actually carry traffic? Most estates carry dead channels. Count the live ones.
- How much custom code is there? Look at transformers, filters, code template libraries, global scripts and anything that calls Java classes or external databases directly. Custom code is where upgrades and migrations cost time.
- Which extensions do you depend on? Any commercial extension, and anything custom-built against the engine's APIs, changes the picture for a fork migration.
- What does each channel do for the business? A lab results feed and a nightly reporting extract do not carry the same risk.
- Who knows how it works? If the answer is one person, that risk exists on every path and should be fixed on every path.
- What does the rest of your roadmap look like? FHIR obligations, EHR changes and acquisitions all touch the engine.
With those answers, the comparison is concrete: effort to upgrade, effort to migrate, license cost over a sensible horizon, and the cost of holding (patching, staff risk, eventual forced change).
A note on timing
There is no deadline built into holding, which is why estates drift. A reasonable plan is to do the inventory now, decide within a quarter, and, if you hold, write down the date you will revisit it.
Where we come in
Our Channel Estate and License-Exposure Assessment produces exactly that inventory and a written recommendation with effort ranges. We are not a reseller and have no stake in which path you take. If the answer is to buy the license and fix the channels, we will tell you that.
Sources: NextGen Healthcare's March 2025 announcement as reported on the Open Integrator community forum; the Open Integration Engine project announcement, April 2025. Licensing terms change, so read the current terms from NextGen Healthcare and each fork project before committing.
Chartferry is independent and is not affiliated with, endorsed by, or a reseller of NextGen Healthcare. Mirth Connect is a trademark of NextGen Healthcare, Inc.