4.2 Previous Mammograms
Key Takeaways
- Comparison with prior mammograms helps the interpreter tell stable findings from true change and is a major reason unnecessary screening recalls (false positives) fall.
- Review priors before positioning so you can match pectoralis muscle, inframammary fold, implant status, clips, laterality, and prior extra views.
- Under 21 CFR 900.12(c)(4), original mammograms and reports are kept at least 5 years, at least 10 years if that patient has no additional mammograms at the facility, or longer if state or local law requires.
- On request, the holding facility must transfer original-modality images (not digitized hardcopy) and copies of reports within 15 calendar days; digital exams released for final interpretation must be provided electronically.
- If priors are elsewhere, acquire today's images, request electronic originals, import any outside CD into PACS with identity checks, and flag the study so interpretation can proceed or be held.
Why Prior Mammograms Are Part of Patient Care
ARRT Patient Care item B.4 is short on the outline and large in clinic: importance of having prior images available and review prior to the exam. Comparison is how breast imaging stays specific. Many densities that look new are old. Many true cancers declare themselves as change. Without the old study, the interpreter has to treat every finding as potentially new, which drives false-positive recalls, extra views, biopsies, and anxiety. With the old study, a stable scar, a stable benign calcification cluster, or a stable intramammary node can be left alone.
You are the person who discovers that the priors exist, that they live on a compact disc in a purse, or that they are sitting in another facility's archive. That work is Patient Care, not clerical extra credit.
Review Priors Before You Position
Review means look at the prior images and the prior report before you set compression, not after the patient has left. You are matching technique and anatomy, not rendering a BI-RADS assessment.
Use priors to learn:
- How steep the prior MLO was and how much pectoralis muscle was included
- Whether the inframammary fold and posterior nipple line were achieved
- Implant presence, clips, skin markers, and laterality quirks
- Whether extra views (spot compression, rolled views, exaggerated craniocaudal) were needed last time
- Breast thickness and exposure so today's automatic exposure control is not a surprise
- Which facility and which modality produced the last study (full-field digital versus digital breast tomosynthesis)
If the prior MLO missed the axilla on a patient who now has an axillary sentinel-node scar, you still try for tissue, but you will not be shocked when mobility is limited. If the prior shows implants, you walk into the room with an implant protocol instead of discovering the implant under the first paddle.
Do not call a glance at a paper printout of a digital tomosynthesis stack a complete review. Printed copies are not the original digital modality.
Comparison Cuts False Positives and Shows Change
Screening programs recall patients for findings that are not cancer. That recall rate is the false-positive burden when the workup is negative. Comparison is the cheapest specificity tool in the department.
| Comparison result | What it usually means | What it prevents |
|---|---|---|
| Finding unchanged across years | Likely benign stability | Unnecessary callback and biopsy |
| Finding new or enlarging | True change until proven otherwise | False reassurance |
| Finding only on today's poorer position | Technique difference | Calling a positioning miss a mass |
| No priors at all | Every density is new by default | Inflated recall rate |
Stability is not a lifetime guarantee, but it is powerful when the images are comparable. Change is why you also should not crop today's posterior tissue just because last year's technologist did. Match the good priors; do not copy the inadequate ones.
A practical exam trap is thinking priors exist only to catch cancer. They also prevent overcalling. A postoperative distortion that has looked the same for four years is not a new architectural distortion. Without year-minus-one, it looks like one.
When Priors Are at Another Facility
Ask every patient where the last mammogram was done. If it was not you, start a request the same visit: patient authorization, destination, and whether they brought a disc. FDA 21 CFR 900.12(c)(4) requires the holding facility, on request by or on behalf of the patient, to transfer original mammograms and copies of reports to a medical institution, the patient's provider, or the patient, within 15 calendar days of receiving that request. Copies of digital studies released for final interpretation must be provided electronically. Any fee charged to the patient for that transfer or copy service may not exceed the documented costs of providing it.
That clock binds the sending facility. Your job at the receiving facility is to request promptly, document the request, and not strand the patient in the dressing room while arguing about mail.
Proceed versus delay interpretation are different decisions.
- Acquisition: Perform today's mammogram unless a specific diagnostic comparison protocol requires the old images in the room (uncommon for routine screening). Do not cancel a screening patient because the disc is in the mail.
- Interpretation: The radiologist may issue a report noting that comparison studies were unavailable, or hold the final interpretation until the originals arrive. Flag the study. Do not pretend comparison occurred.
Never tell a patient that MQSA forbids imaging without priors. It does not. MQSA forbids you from losing the originals you already have and from substituting digitized hardcopy for those originals.
MQSA Recordkeeping You Can Quote
Under 21 CFR 900.12(c)(4), each facility that performs mammograms must maintain the original mammograms and mammography reports in a permanent medical record for the longest of:
- a period of not less than 5 years,
- a period of not less than 10 years if no additional mammograms of that patient are performed at the facility, or
- any longer period mandated by state or local law.
Facilities must have policies to minimize loss. Originals must be kept in retrievable form in the mammographic modality in which they were produced. They cannot be produced by copying or digitizing hardcopy originals. Before a facility closes or ceases mammography, it must arrange continued access for the remainder of those retention periods.
Those numbers are exam-relevant because they explain why last year's images should still exist and why a we-shred-after-two-years story is not the federal standard. If a state requires longer retention, the longest period wins.
| Rule in 21 CFR 900.12(c)(4) | Practical meaning for the mammographer |
|---|---|
| Keep originals at least 5 years | Last year's study should still be retrievable here |
| Keep at least 10 years if the patient never returned | One-time visitors are not a 90-day file |
| Longer if state or local law says so | Know your state's overlay |
| Original modality, not digitized film | Do not scan printed films of digital exams and call them originals |
| Transfer within 15 calendar days of request | Start the request the day the patient is in front of you |
| Digital copies for final read must be electronic | A paper printout is not an interpretation original |
Digital Priors, Printed Film, Outside CDs, and PACS Import
Digital full-field mammography and digital breast tomosynthesis (DBT) are native electronic studies. Comparison should happen on a workstation using the electronic files, including DBT slices when the prior was tomosynthesis. A printed film of a digital exam is a photograph of a dataset. It is useful as a last-ditch preview. It is not the original modality. MQSA is explicit that examinations presented for interpretation must be in the original mammographic modality and must not consist of digital images produced by copying or digitizing hardcopy original images.
If a patient hands you an outside CD:
- Confirm it is that patient's study (name, date of birth, exam date).
- Import through the facility picture archiving and communication system (PACS) or import gateway so the radiologist can hang priors next to today.
- Check laterality and view labels; foreign sites do not always use your mnemonic.
- Prefer DICOM. A folder of JPEGs is not a mammography original.
- For DBT, import the stack, not only a synthesized two-dimensional image, when the stack is what was acquired.
If import fails, document the failure, still complete today's exam, and keep requesting a proper electronic transfer. Do not digitize printed copies of digital mammograms and label them originals. That is the practice 21 CFR 900.12 is written to stop.
Comparison with prior mammograms primarily helps the interpreting physician do which of the following?
Under 21 CFR 900.12(c)(4), a mammography facility must keep original mammograms and reports for the longest of which periods?
Priors are at another facility and are not yet in PACS. What is the best technologist action?