Summary
There is no single veterinary rule for how long to keep radiographs. Each state or country sets its own period, and many count it from the patient's last visit rather than the day the image was taken. Whatever the period, the archive has to hand back a complete study, still matched to the right patient, when someone asks for it. A change of system is where that usually gets tested.
How long do you have to keep veterinary radiographs?
Retention is set by the veterinary board or regulator where you practise. The US has no federal standard, and state rules differ both in length and in when the clock starts.
New Jersey shows how these rules tend to be written. Its regulation, N.J.A.C. 13:44-4.9, requires records and radiographs to be kept for five years from the patient's last visit. When the record shows the patient has died, the period is three years from the last entry. Other states set different periods, and some count from the last entry in the record instead of the last visit.
Australia also varies by state and territory. The Australian Veterinary Association points to seven years, the period in the ASAV hospital accreditation standards, and is clear that the legal requirement is whatever the practice's own jurisdiction sets.
Three practical consequences follow from that:
- Where the clock runs from the last visit, a radiograph of a patient you still see can need keeping long after the headline period has passed since it was taken.
- Seven years is a common working policy. It is not a universal legal minimum, so confirm the current rule with your own board.
- Rules like New Jersey's name radiographs explicitly, so the retention period covers the DICOM files and not only the written notes.
What does an archive have to do for that whole period?
Writing a file to storage is the easy part. The test comes years later, when someone asks for one specific study for a referral, a second opinion or a complaint. Three things have to be true on that day.
The study is complete. Every series and every image that left the modality reached the archive. A CT that archived 400 of its 412 slices still looks normal in a study list. The gap only shows when someone scrolls through it.
The study still matches the patient. A DICOM file carries its own identification: patient ID, study and series instance UIDs, modality and dates. If those were thin or malformed at acquisition, the study can open in the system that stored it and still fail to match a patient record anywhere else. A 2018 survey of referral studies in Germany found that only 5.1% fully complied with the recommended veterinary identification tags. Our guide to veterinary DICOM conformance covers what that means and how to check your own archive.
The study can leave. If you cannot export your archive in standard DICOM, you can only read your own history through one vendor. That becomes a problem the day you want to change systems or send a full history to a specialist.
Questions to put to any archive vendor
These are worth asking of whatever system holds your images today, and of anything you are considering:
- How do you confirm a study arrived complete, and what happens when it did not?
- Are stored files checked again over time, for example against a checksum, or only when they arrive?
- Where are copies held, and how many are there?
- What happens when a study reaches the end of its retention period? Is anything deleted automatically?
- How do I export the whole archive in standard DICOM, reports included?
- If a file fails on import, how am I told, and who fixes it?
A vendor who answers these with a mechanism, not a reassurance, has thought about the archive as a record rather than a disk.
What goes wrong when you move an archive?
Migration is where the quality of an archive becomes visible. Studies stored with weak identification often open normally in the system that created them, because that system learned to tolerate its own data. A new system that applies the standard more strictly rejects them, or imports them without a patient match. These steps catch most of that before it becomes a gap in the record:
- Get a full export before you sign. Ask for standard DICOM, not a proprietary wrapper, and ask for the reports as well. Reports are often stored apart from the images, and a migration that moves images without reports loses half the record.
- Validate a sample. Run part of the export through a free DICOM validation tool such as dciodvfy. It shows which studies have identification problems while they are still easy to fix.
- Agree how failures are handled. Failed files should be reported one by one, with a named owner for the fix. A failed batch with no detail is not a migration plan.
- Check that original dates and identifiers survive. A study re-stamped with its import date breaks your retention calculations and confuses prior matching when a radiologist pulls up earlier studies.
- Set a cutoff if the volume is very large. Moving decades of images can take a long time. Agreeing a date for image data, while report text transfers in full, keeps the switch on schedule.
How does NewLumen store, keep and import studies?
NewLumen is Asteris's cloud-native veterinary DICOM platform. Image management, viewing, reporting, scheduling and teleconsultation all run in the browser, with no workstation install. Here is how it handles the points above.
Archival streams as images arrive. A radiologist can open and report a study while it is still archiving. When the remaining images land, the report does not need an addendum.
Two ways to deploy. Modalities can send straight to the cloud, with no server on site. For practices doing heavy cross-sectional work, or whose IT team wants images on the local network, the edge node is a lightweight Windows install. It caches studies locally, so images are viewable in sub-seconds as they stream in. It spools everything to the cloud in the background, and keeps local access going through an internet or power outage.
Nothing is deleted without consent. When studies reach the retention limit, the practice is notified. It can extend, export, or continue at a long-term storage rate.
Imports run in the background with per-file error handling. A file that fails is reported on its own, so one bad study does not hold up the rest of the history.
Migration is part of onboarding. Studies and reports both move, and historic reports are fully indexed and searchable afterwards. Where imaging volume is very large, we agree a cutoff date for image data, and report text transfers in full.
Studies can go out by DICOM send or by a share link with an expiry date. What NewLumen does not offer today is DICOM query and retrieve for external viewers, so another system cannot pull studies from it directly yet.
To see how this fits a cross-sectional workload, the NewLumen product page covers the viewer and deployment options in more detail. Our cloud vs on-premise guide covers where storage should live.
FAQ
How long do veterinary practices have to keep radiographs?
It depends on where you practise. Each US state sets its own period, and there is no federal standard. New Jersey, for example, requires records and radiographs to be kept for five years from the patient's last visit. In Australia the Australian Veterinary Association points to seven years, while noting that each state or territory sets the legal requirement. Confirm the current rule with your own veterinary board.
Does the retention period start on the day the radiograph is taken?
Often it does not. Many rules count from the patient's last visit or the last entry in the record. Under a last-visit rule, a radiograph of a patient you still see has to be kept for as long as the patient keeps coming in, plus the retention period. Check which starting point your jurisdiction uses before you set an archive policy.
Why does a stored DICOM study become hard to use years later?
Usually because of its identification, not its pixels. If the patient ID, study and series UIDs or dates were incomplete when the image was acquired, the study may open in the system that stored it but fail to match a patient record anywhere else. A 2018 survey of German referral studies found only 5.1% fully met the recommended veterinary identification tags.
What should I ask for before leaving my current PACS?
A full export in standard DICOM, with reports included, before you sign with a new vendor. Validate a sample with a free DICOM validation tool, agree how failed files will be reported and fixed, and confirm that original study dates and patient identifiers survive the move. For very large archives, agree a cutoff date for image data while report text transfers in full.
What happens to studies in NewLumen at the end of the retention period?
Nothing is deleted without the practice's consent. When studies reach the retention limit, NewLumen notifies the practice, which can then extend retention, export the studies, or keep them at a long-term storage rate. Studies can also be sent out at any time by DICOM send, or shared through a link with an expiry date.
Sources
- New Jersey Administrative Code, N.J.A.C. 13:44-4.9(e), patient records and radiographs.
- Australian Veterinary Association, policy on retention of medical records and diagnostic images, citing the ASAV Manual of Hospital Standards and Accreditation 2019.
- Brühschwein A, Klever J, Wilkinson T Jr, Meyer-Lindenberg A. DICOM Standard Conformance in Veterinary Medicine in Germany: a Survey of Imaging Studies in Referral Cases. J Digit Imaging. 2018;31(1):13-18.