How to Dictate a Screening Mammogram: MQSA, BI-RADS Verbatim, and the Discipline of the Negative

The screening mammogram is the only report most residents will ever dictate whose required contents are written into federal regulation. MQSA — the Mammography Quality Standards Act — is not a style guide; it is the reason every mammography report in the country, from a rural two-view unit to an academic tomosynthesis program, contains the same skeleton. Learn the skeleton once and the dictation takes forty-five seconds. Freelance, and you produce reports that fail audit.

What the report must contain

Every screening report carries, at minimum:

• Patient identity and exam date — handled by the RIS, but the report is the legal document, so verify the header you're signing under.
• A breast density statement. The four BI-RADS categories, and you commit to one: almost entirely fatty; scattered areas of fibroglandular density; heterogeneously dense, which may obscure small masses; or extremely dense, which lowers the sensitivity of mammography. The qualifying clauses on the dense categories are part of the standard language — they are the sentence that tells the referring clinician and the patient why a negative mammogram is not a guarantee, and federal rules now require density notification to reach the patient directly. Don't trim them.
• The final assessment category, verbatim. "BI-RADS 1: Negative." "BI-RADS 2: Benign." "BI-RADS 0: Incomplete — need additional imaging evaluation." The number and the standard phrase, exactly. This is the field the facility's outcome audit is built on, the field the lay letter to the patient is generated from, and the field a tracking system reads. A paraphrase like "no evidence of malignancy" with no category is not a screening report; it is a liability with a header.
• Comparison. Name the priors and their dates. Stability over time is the single most powerful benignity argument in screening, and "no prior studies available for comparison" is itself a required statement when true.

A screening exam ends in 0, 1, or 2. The suspicious categories — 4 and 5 — belong to the diagnostic workup, after the callback imaging is done. The screener's job is a binary: dismiss or recall. Resist the urge to pre-stage suspicion from four views.

Why the category is quoted, not paraphrased

The verbatim rule feels pedantic until you follow the category downstream. MQSA requires the facility to send every patient a lay-language summary of her result, and that letter is generated from the assessment category — a report with a mangled or missing category produces a mangled or missing letter. The category also feeds the facility's medical outcomes audit: recall rate, cancer detection rate, and positive predictive value are all computed by counting categories, so a "negative, but attention to the left breast on next year's study" corrupts the statistics in both directions at once. And when a patient moves and her films follow her, the next radiologist reconstructs your assessment from one field. The prose can be yours; the category belongs to the system. One category per report, bilateral exams carrying the more severe side's assessment, and no local dialects — "BI-RADS 1-2" is not a category, and neither is "essentially negative."

The distinction between 1 and 2 is worth one deliberate beat: negative means nothing to describe at all; benign means you described something — a stable fibroadenoma-pattern calcification, an intramammary node, a secretory calcification — and are affirmatively dismissing it. Both end the encounter until next year. Pick the one that matches what your findings section actually said.

The discipline of the negative

Most screens are negative, and the negative report is where dictation discipline is most visible. Two rules:

Say less. The negative screen does not inventory every stable calcification, every lymph node with a fatty hilum, every island of asymmetric fibroglandular tissue. Cataloguing benign findings trains the referring clinician to skim, and the day your report finally contains a sentence that matters, it will be skimmed too. "Scattered benign-appearing calcifications, stable" is one line, category 2, done.

But say it the same way every time. The negative body is a fixed paragraph: composition, no dominant mass, no suspicious calcifications, no architectural distortion, comparison statement. A fixed paragraph is auditable, fast, and — critically — any deviation from it is a signal. Our normal screening mammogram template is exactly this paragraph; the template library keeps the same philosophy across modalities.

The hardest skill in screening is not finding cancer. It is declining to recall the stable, the symmetric, and the benign — because every recall has a cost in anxiety, biopsies, and program statistics — while recalling the one subtle asymmetry that is new. The report language reflects that economy: a negative is short and unhedged; a recall is specific and unapologetic.

Recall wording: BI-RADS 0 done properly

When you do recall, the report has three jobs — say what, say where, say what happens next:

• What: use the lexicon noun. Mass, asymmetry, focal asymmetry, developing asymmetry, architectural distortion, or calcifications (with their distribution — grouped, linear, segmental). "Density" is not a finding noun; it's what the whole breast has.
• Where: laterality, quadrant or clock-face position, and depth (anterior, middle, or posterior third). The diagnostic technologist will set up spot compression from your sentence. "Left breast upper outer quadrant, middle depth, seen only on the MLO view" is a workup plan; "possible left asymmetry" is a shrug.
• Next: the specific recommendation — "recall for diagnostic mammography with spot compression views; ultrasound as needed." If the finding is seen on one view only, say so, because that changes the callback protocol.

And the category line reads: "BI-RADS 0: Incomplete — need additional imaging evaluation." Not "probably benign, recommend additional views." Category 3 is a diagnostic-exam decision made after a complete workup, not a screening hedge.

The phrases attendings correct

• "No significant abnormality." The word is negative, and it has a category number attached.
• "Stable when compared to prior" without naming the prior's date — the audit trail needs the date, and so does the next radiologist.
• "Dense breasts" as the whole density statement. Pick one of the four categories and use its language.
• Hedged recalls — "slightly asymmetric tissue, probably summation, attention on next screen." If it deserved that sentence, it deserved a callback. Screening is binary; write like it.

A model report

EXAMINATION: Screening digital mammography with
tomosynthesis, bilateral
HISTORY: 52-year-old female, annual screening.
Asymptomatic.
COMPARISON: Screening mammograms of 2025-10-14 and
2024-09-30.
FINDINGS:
The breasts are heterogeneously dense, which may obscure
small masses.
There is no dominant mass, suspicious calcification, or
architectural distortion in either breast. Scattered
benign-appearing calcifications are stable. No new or
enlarging asymmetry when compared with prior
examinations.
IMPRESSION:
1. Negative screening mammogram. No mammographic
   evidence of malignancy.
2. Breast composition: heterogeneously dense.
BI-RADS 1: Negative.
Routine annual screening mammography is recommended.

Nine lines of findings, one unambiguous category, dates on the comparison. That is the whole craft, repeated a hundred times a day without drift — which is precisely the kind of discipline reporting software should hold steady for you. Watch a screening case draft itself in the live demo — the density statement, the verbatim category, and the recall logic are all enforced, so the report you sign is the report the audit expects.