It is not every day a boy band turns up in a serious conversation about healthcare IT. But at ViVE 2026, Mike Prokic, Chief IT Strategy Officer at Trinity Health, reached back to a nineties breakup anthem to describe what actually makes legacy application decommissioning hard at scale:
"In the nineties, there was this song called 'The Hardest Part of Breaking Up Is Getting Back Your Stuff.' The point is, if you've ever broken up with a vendor, that's when things get very contentious. Data all of a sudden becomes very proprietary, proprietary formats, and resources aren't available."
The observation is accurate. The moment a health system moves to retire an application, it often discovers that getting its own data back is not as simple as it should be. Information essential to patient care, compliance, operations, and reporting gets locked inside proprietary formats, held hostage by limited vendor resources, or tied to contracts the organization no longer wants to fund. The health system wants a clean break-up, but it needs to get its stuff (data) back.
It usually happens gradually. Most health systems do not set out to carry hundreds of aging, redundant or lightly used systems. The portfolio grows through mergers, acquisitions, EHR implementations, ERP rollouts and departmental buying decisions. An Epic install here, a Workday rollout there, and suddenly the organization is maintaining a long tail of applications that still hold critical data, even if those systems no longer reflect how the enterprise operates.
Gartner documents the scale of the problem in its research on application rationalization: nearly half of application and software engineering leaders cite maintaining legacy applications as one of their top challenges when implementing innovative technology. The backlog grows because the organizational incentive is always to implement, never to retire.
Every application in the portfolio carries a supplier, a contract, a renewal date, a data model, a retention obligation, and a security surface. That is how technical debt becomes an unnecessary operational expense. The application may no longer be central to the business, but the data inside it still needed. Clinicians need it for patient history. Health information management needs it for release-of-information workflows. Finance, compliance, legal, and operations need it for audits, reporting, and historical reference.
Legacy decommissioning becomes more complicated than deciding what to turn off. By the time a health system is ready to leave a vendor, it must also navigate contract terms, termination windows, data ownership, extraction requirements, archival standards, user access needs and format conversion. If that work begins when the renewal notice arrives, the organization is already late.
So, health systems face two bad options: keep paying maintenance on a system hardly anyone uses or attempt the extraction and find out exactly how cooperative the soon-to-be-former vendor intends to be.
Gartner has a term for what gets left behind: zombie applications. Not fully alive, not fully dead — systems that linger in the portfolio consuming budget, requiring security patching and infrastructure support, while providing little to no active business value. Gartner research on the application undertaker notes that decommissioning consistently fails when it is bundled into the replacement project rather than treated as a separate, funded initiative with its own owner and governance.
Many decommissioning efforts stall right there. Teams extract the structured data, declare progress, and quietly leave the hard work behind: the imaging, the DICOM files, the scanned documents, the records trapped in formats nobody wants to touch. The old system never fully dies. The contract never fully ends. The savings are never fully realized.
There are not fifty ways to leave a legacy vendor cleanly. At scale, one path works: build the discipline to rationalize, decommission, and actively archive as a standing enterprise capability.
That starts with prioritization. Gartner's application rationalization research is consistent on this point: 20% of applications typically account for 80% of the cost, complexity, and risk in a portfolio. Health systems do not need to inventory everything before they start. They need to identify where the financial exposure is highest and move on those systems first.
It also requires separating the work. Application replacement and application decommissioning are two distinct initiatives with distinct ownership. Without that separation, the decommissioning work consistently loses to the implementation team's priorities. Health systems need a clear owner, a funded program, and shared governance across finance, IT, legal, compliance, security, HIM, analytics and operations — with visibility into contract terms, renewal dates, termination rights, data requirements and archival needs before the next exit window arrives.
They also need an operating model built for speed. The assembly-line approach — moving applications from rationalization and decision-making through data acquisition, archival and decommissioning in sequential waves — is what turns a one-time project into a permanent enterprise function. That structure matters because every delayed extraction creates risk. Every missed termination window gives the old vendor more leverage. Every one-off negotiation slows the next wave.
The financial model matters too. Done correctly, application rationalization and active archiving are not IT work. They are a cost-takeout strategy. Health systems can use capital budgets to fund eligible implementation work while creating permanent operating expense reduction by retiring redundant systems, eliminating licensing and maintenance costs, lowering infrastructure burden and shrinking the attack surface. The CapEx/OpEx structure makes this fundable in ways a traditional IT project is not.
The returns are documented. A Gartner case study on large-scale decommissioning in healthcare found that one major not-for-profit integrated delivery network archived data from over 540 applications and systems and achieved approximately $68 million in annual operating expense savings — eliminating recurring licensing, support, maintenance, infrastructure and staffing costs at a rate of 20 to 45 applications per month. The program established a patient-centric repository that also enables population health analytics and AI model training on historical longitudinal data that was previously inaccessible.
The value does not stop at cost takeout. Retiring legacy applications shrinks the number of systems that require patching, monitoring, and defense, reducing cybersecurity exposure. It accelerates M&A integration by shortening the time and cost associated with transition service agreements and redundant platforms. And it creates a cleaner foundation for modernization, analytics and AI by making historical data more accessible, governed and usable.
That is why the breakup metaphor holds. The goal is not simply to end the vendor relationship. The goal is to have a clean “breakup with the data intact, the security risk reduced and the organization free to modernize without continuing to fund systems it no longer needs.
In healthcare IT, getting your stuff back is not the punchline. It is the strategy.