PowerScribe 360 Migration Checklist and 90 Day Plan

The short answer: A PowerScribe 360 migration takes most radiology groups three to six months of real calendar time, and the work splits into four phases: inventory what you actually own and use, decide the destination before you take a quote, do the interface and template work, then pilot with a small group before cutover. Annual renewals and maintenance for PowerScribe 360 ended on August 31, 2026, and full support ends on August 31, 2027, so a group starting now still has runway. A group starting in mid 2027 does not.

Most radiology practices have not changed reporting systems in a decade. That is the real difficulty here. The technical work is well understood and vendors do it routinely, but the institutional knowledge of how your last transition went has usually walked out the door, and the templates and macros your readers depend on were built up by people who may no longer work there. This checklist is written for the person who has just been handed the project.

What are the PowerScribe 360 end of life dates?

Microsoft sent end of life letters to PowerScribe 360 customers in early 2026. Two dates matter and they do different jobs in your plan.

The gap between those two dates is the useful part. You are not being cut off tomorrow, but you also cannot buy your way into another year of maintenance, which is the option groups normally reach for. We covered the background and the vendor context in more detail in our piece on what the PowerScribe 360 end of life means for radiology groups.

Phase 0, weeks 1 and 2: find out what you actually own

Nobody starts a migration knowing this, and every group that skips it discovers something expensive in week nine. Get the following on one page before you talk to any vendor.

• Your contract. License type, seat count, current maintenance end date, any auto renewal language, and who signed it. Finance often has a different version of this than IT does.
• Your version and infrastructure. Which PowerScribe 360 release you are on, what servers it runs on, what those servers also do, and who patches them.
• Every interface. The HL7 order feed, the results interface back to the RIS or EHR, the PACS integration, any downstream feed to a data warehouse, peer review tool, billing system or critical results notification system. Interfaces are where migration timelines go to die, and there are always more of them than the diagram shows.
• Your template library, with usage counts. Not how many templates exist. How many were used in the last ninety days, and by whom. Groups routinely find that 400 templates in the system resolve to 60 in genuine daily use, which turns a terrifying migration task into a manageable one.
• AutoText, macros and personal shortcuts. Including the ones individual radiologists built for themselves and never told anyone about. These are the single most common cause of post cutover unhappiness.
• Who reads, from where. On site, home workstations, teleradiology partners, night coverage. Each one is a different network path to validate.

It is worth treating this phase the way you would any other operational readiness review that scores where the group actually stands on process and systems, rather than as an IT inventory. The gaps it surfaces are usually organizational, not technical: an interface nobody owns, a template library nobody has pruned since 2019, a night coverage arrangement that was never documented.

Phase 1, weeks 3 to 5: decide the destination before you price it

The default assumption is that a PowerScribe 360 site moves to PowerScribe One. That is Microsoft's recommended path, and for plenty of groups it is the right one. But taking a quote before you have decided the destination means you evaluate one option against nothing.

There are three realistic destinations, and they are worth naming explicitly:

1. PowerScribe One. The cloud subscription successor. Familiar dictation, cloud speech recognition, AI-assisted impressions, and a migration path Microsoft says reuses your existing 360 workflows and content. The trade is that a perpetual license becomes a renewing subscription and uptime becomes a vendor responsibility. Our PowerScribe One vs 360 comparison sets the two out side by side, including where the older product still wins.
2. A different speech-based reporting platform. Several vendors have built their sales motion around this exact moment. You will get attention and sharp pricing right now, which is worth using.
3. A reporting layer that does more than dictation. The end of life is the cheapest moment your group will ever get to ask whether transcribing your own speech is still the best way to produce a report. Drafting the structured report into your template before the radiologist sits down changes the job from construction to editing. Prioritizing the worklist by suspected urgency changes which study gets opened next. Neither of those is a dictation feature.

Whatever you shortlist, send every vendor the same written questions in the same email on the same day. Our vendor evaluation question list is built for that, and the answers are far more comparable than three separate sales calls will ever be.

On pricing: ask for the fully loaded first year total, including interface work, template migration, training and any professional services, and ask separately what the subscription does at renewal in year three. Almost nobody in this category publishes list pricing, so the only useful comparison is total cost against total cost. We break the components down in how much radiology AI costs.

Phase 2, weeks 6 to 9: interfaces and content

This is the longest phase and the one that slips. Two workstreams run in parallel.

The interface work. Every integration from your Phase 0 list needs an owner, a test case and a sign off. Order feed in, results out, PACS launch context, and each downstream consumer. The specific thing to test is not whether a message passes, but whether the signed report lands correctly in the EHR with the right accession, the right addendum behavior and the right handling of a corrected report. Addenda and corrections are where interface bugs hide, because they are rare enough not to appear in a happy path test.

The content work. Migrate the 60 templates that are actually used, not the 400 that exist. For each one, run a real signed report from the last quarter through the migrated template and compare the output line by line. Have the radiologist who owns that template do the comparison, not the project manager. Personal AutoText and macros need an owner too: give every reader a scheduled thirty minute session to move their own shortcuts, because the alternative is finding out at cutover.

Do not let anyone tell you the template migration is automatic. Microsoft describes a path that leverages existing PowerScribe 360 content rather than rebuilding from scratch, which is genuinely helpful, but "leverages" is not "verified against your signed reports". That verification is your job.

Phase 3, weeks 10 to 13: pilot, cutover, decommission

Run a pilot with three to five radiologists across different subspecialties for at least two full weeks, including a weekend and a night shift. The people to pick are not the enthusiasts. Pick one enthusiast, one skeptic and one high volume reader, because the high volume reader will find the friction the other two tolerate.

Set the exit criteria before the pilot starts. Reasonable ones: no unresolved interface defect, median report turnaround within a defined margin of your pre migration baseline, and every pilot reader willing to sign in the new system without a workaround. If you have not measured your current turnaround time before the pilot, you have no baseline and the pilot cannot fail, which sounds convenient and is not.

Cut over in waves rather than all at once. PowerScribe 360 and PowerScribe One can operate side by side during the migration period, and if you move to a different platform entirely you should insist on the same overlap. Move one site or one subspecialty per wave, hold a week between waves, and keep a documented rollback for each.

Then decommission deliberately. Confirm your report archive is accessible independently of the retired system, confirm retention obligations are met, formally end the old contract, and reclaim the server capacity. Groups routinely pay for a year of infrastructure nobody has logged into.

How long does a PowerScribe migration take?

Three to six months of calendar time for a typical group, of which perhaps six to eight weeks is actual technical work. The rest is scheduling: getting radiologists into template review sessions, waiting for interface engineering resources, running the pilot long enough to mean something, and holding between cutover waves. Large multi site health systems should plan for longer. The calendar, not the engineering, is what sets the date.

What happens if we do nothing?

Full support for PowerScribe 360 ends August 31, 2027. After that the software does not stop working, but you have no vendor to call when it breaks, no security patches for the stack it runs on, and no supported path for an operating system or database upgrade underneath it. The practical risk is not a dramatic failure on the deadline. It is a Windows update or a hardware failure eighteen months later with nobody obligated to help, at which point you are running an unplanned migration during an outage instead of a planned one on a schedule.

Do we need to rebuild our templates from scratch?

Generally no. Microsoft's stated migration path reuses existing PowerScribe 360 workflows and content, and other vendors bidding for this business will import templates as part of winning it. What you do need to do is verify the migrated versions against real signed reports, and use the moment to retire the templates nobody has opened in a year. A migration is the only time a group will willingly prune a template library, so it is worth the two extra weeks.

Who should be on the migration team?

Five roles, and they can be four people. A radiologist lead with genuine credibility among the readers, because template and workflow decisions need clinical authority behind them. An interface or integration engineer who owns every HL7 and DICOM connection. A PACS administrator. Someone from finance or contracting who owns the commercial terms and the renewal language. And a project manager who holds the schedule. The failure mode is a project run entirely by IT, which produces a technically correct migration that radiologists resent.

The short version

Inventory before you shop. Decide the destination before you take a quote. Count the templates that are genuinely in use. Own every interface. Pilot with a skeptic and a high volume reader. Cut over in waves with a rollback. And use the fact that you are spending this effort anyway to ask whether the reporting layer you are migrating to should do more than transcribe your voice, because you will not get a cheaper moment to ask for another decade.