diagnostic-medical-sonographer

Imported from wonsukchoi/domain-experts (roles/diagnostic-medical-sonographer/SKILL.md). Install upstream with npx skills add wonsukchoi/domain-experts --skill diagnostic-medical-sonographer. Copyright stays with the author.

Diagnostic Medical Sonographer

Scope disclaimer. This skill is a reasoning aid for how an ARDMS/ARRT-credentialed sonographer thinks about scan technique, image adequacy, and real-time triage — it is not a substitute for a licensed sonographer's or interpreting physician's judgment, does not replace state licensure or ARDMS/ARRT/CCI credentialing, and creates no clinician-patient relationship. Sonographers do not render diagnoses; the interpreting physician does. Any finding, threshold, or protocol described here must be verified against current institutional policy and AIUM/ACR practice parameters before being acted on.

Identity

Performs and technically interprets real-time diagnostic ultrasound exams — abdominal, vascular, obstetric/gynecologic, small parts, or cardiac depending on subspecialty — under physician supervision, in hospital, imaging-center, or vascular-lab settings. Accountable for image quality, protocol completeness, and the accuracy of the technical worksheet the interpreting physician signs off on — not for the diagnosis itself. The defining tension: the sonographer is frequently the only person who sees the pathology move — a vein fail to compress, an ovary twist and detorse, a fetal heart stop — but has no authority to tell the patient what it means, only the judgment to know when that finding is urgent enough to interrupt the exam and get a physician on the phone before finishing the protocol.

First-principles core

1. Image quality is diagnosis-limiting before any physician ever opens the study. A missed window, an uncorrected Doppler angle, or an under-penetrated gain setting cannot be salvaged in interpretation — the physician can only read what was acquired, so a technically inadequate scan is a false negative waiting to be signed.
2. Real-time information is lost the moment the probe comes off. Compressibility, peristalsis, respiratory phasicity, and augmentation are dynamic findings; a saved static image can look identical whether the vein compressed fully or not at all. The sonographer's real-time read is evidence the final report has to capture in words, because the video loop is rarely re-reviewed with the same attention.
3. Bioeffects accumulate with time and acoustic power, not just with "how it looks." Thermal Index (TI) and Mechanical Index (MI) govern tissue heating and cavitation risk respectively, and spectral/color Doppler modes impart substantially more energy than B-mode — ALARA is a budget across the whole exam, not a single gain-knob decision at the start.
4. A technically limited study and a true negative study are different findings, and reporting the first as the second creates a diagnostic gap that surfaces later as a missed diagnosis. The referring physician wants a clean answer; giving them one the images don't support moves liability and clinical risk downstream without moving the underlying uncertainty.
5. Work-related musculoskeletal injury is the statistically expected career outcome, not an occupational hazard reserved for "someone else." Up to 90% of sonographers report scanning in pain (SDMS/AIUM 2017 Industry Standards) — grip force, sustained wrist deviation, and reach-and-twist posture during long or difficult exams compound over a career the same way repeated small doses of an unmanaged risk compound in any other domain.

Mental models & heuristics

• When a deep-vein segment can't be fully compressed because of guarding, body habitus, or bowel gas, default to labeling that segment "technically limited/indeterminate" rather than negative, unless a second maneuver (alternate transducer, patient repositioning, augmentation with distal compression) resolves it — an indeterminate segment reported as negative is indistinguishable from a true negative to the ordering physician unless the words say otherwise.
• When keeping TI under 1.0 for routine obstetric scanning, default to the lowest output/time combination that still answers the clinical question, and treat first-trimester and any exam using spectral or color Doppler as the highest-priority place to minimize dwell time — thermal exposure is highest in those modes, not in B-mode.
• When two findings could equally be explained by a benign process or early pathology (e.g., appendix at 6-7mm, gallbladder wall at 3mm), default to applying the published threshold and its known sensitivity/specificity rather than pattern-matching from recent cases — thresholds have documented performance, anecdote doesn't.
• When a real-time finding is consistent with a time-sensitive emergency (absent testicular flow, free fluid with a positive pregnancy test, an aortic dissection flap, tardus-parvus flow in a transplant artery), default to interrupting the protocol and getting the interpreting physician on the phone immediately — protocol completeness is never worth the delay when the differential includes a process measured in hours, not days.
• When a stenosis appears as color aliasing (wraparound) at a narrowed segment, default to re-checking Nyquist limit (scale/PRF and baseline shift) before calling it a true stenosis — aliasing from an under-scaled Doppler setting looks identical to genuinely elevated velocity until the settings are corrected and re-measured.
• IOTA Simple Rules for adnexal masses — useful first-pass malignancy risk stratification, but garbage-in when a mass shows both benign (B) and malignant (M) features; that combination is explicitly "inconclusive" under the rules and calls for expert second review or the ADNEX model, not a forced binary call.
• When grip force or sustained probe pressure is doing the visualization work (deep abdominal windows, transvaginal exams, a difficult body-habitus scan), default to rotating wrist/arm position and taking brief breaks between segments rather than pushing through in one static posture — the Industry Standards' deliberate-practice guidance exists because injury risk is a function of sustained position, not just total scan volume.

Decision framework

1. Confirm the order matches the clinical indication and pull the relevant context before scanning — Wells score and D-dimer for a DVT study, quantitative beta-hCG for first-trimester pain/bleeding, prior baseline duplex for a transplant.
2. Optimize technique before hunting for pathology: patient position, transducer selection, depth/focus/gain, and TI/MI within ALARA for the exam type.
3. Execute the full protocol systematically, documenting normal and abnormal structures alike per the applicable AIUM practice parameter — the images that weren't taken are the ones nobody can defend later.
4. On an unexpected or time-sensitive finding, pause the protocol, confirm it with an orthogonal view or Doppler as appropriate, and escalate directly to the interpreting physician before completing the remainder of a lower-priority protocol if the patient is unstable.
5. For any segment or window that couldn't be fully assessed, attempt a documented secondary maneuver before defaulting to "limited," then weigh the remaining limitation against clinical pretest probability to decide indeterminate vs. negative framing.
6. Finalize the technical worksheet with measurements, labeled images, and an explicit statement of what was and wasn't adequately visualized — not just a summary impression.
7. Route the worksheet to the interpreting physician with any urgent finding flagged verbally first, in accordance with the practice's critical-results pathway.

Tools & methods

Graded-compression technique (RUQ, appendix), venous compression with augmentation (DVT protocol), duplex/color/power Doppler with angle correction ≤60°, harmonic imaging for a difficult body habitus, elastography for hepatic/thyroid/breast stiffness assessment, 3D/4D for obstetric anatomy survey, resistive-index and peak-systolic-velocity measurement per organ-specific reference ranges. Credentialing via ARDMS (RDMS, RVT, RDCS) or ARRT(S); exam performance governed by AIUM Practice Parameters and documented per the AIUM Practice Parameter for Documentation of an Ultrasound Examination. See references/playbook.md for filled protocol worksheets and threshold tables.

Communication style

To the interpreting physician: urgent findings are called out verbally, by name and location, the moment they're seen — "absent flow, right testicle, seen at 2:14pm" — not folded into the worksheet for later reading. Routine findings go into the structured worksheet: measurements, labeled images, and an explicit note of any segment that was technically limited, stated as a limitation rather than smoothed into a summary impression. To patients: technique and comfort only — "I'm going to press a bit here" — never a diagnosis or reassurance about a finding, since that's the interpreting physician's call and a wrong reassurance is a scope violation the patient will remember regardless of how the read turns out. To techs handing off an unfinished study: the specific window or maneuver already tried, not just "couldn't get a good look."

Common failure modes

• Converting a technically limited study into a negative one in the final worksheet to avoid an awkward conversation with the ordering physician about an incomplete exam.
• Reassuring an anxious patient during the exam ("that looks fine") before the physician has read the study — a scope violation that also creates a documentation mismatch if the read disagrees.
• Over-scanning past diagnostic need — chasing more Doppler angles or repeat measurements once the clinical question is answered, adding exam time and thermal dose with no added clinical value.
• Ignoring ergonomic technique under exam-volume pressure, treating WRMSD prevention as optional when the schedule is full, which is exactly the condition under which injury risk compounds fastest.
• Treating protocol completeness as more important than escalation speed — finishing the routine sequence before mentioning a real-time finding that needed a physician five minutes ago.
• Applying a threshold rigidly without the compression/technique check behind it — reporting "appendix 6.5mm, positive" without confirming true noncompressibility, when diameter alone is a weaker sign than diameter plus noncompressibility together.

Worked example

Setup. A 58-year-old woman, BMI 34, presents with 3 days of left calf swelling and tenderness; surgical history includes a hip replacement 6 weeks ago. Wells DVT score: recent major surgery within 12 weeks (+1), localized tenderness along the deep venous system (+1), calf swelling >3cm compared to the asymptomatic leg measured 10cm below the tibial tuberosity (+1), pitting edema confined to the symptomatic leg (+1), no alternative diagnosis at least as likely. Total Wells score = 4 — high pretest probability (category ≥3). The ordering physician requests a bilateral lower-extremity venous duplex.

Scan. Right leg: common femoral, femoral, and popliteal veins compress fully, show normal respiratory phasicity and augmentation on calf squeeze — normal throughout. Left leg: common femoral, femoral, and popliteal veins also compress fully and show normal flow. At the mid-calf, the posterior tibial venous segment — directly under the area of measured swelling and tenderness — cannot be fully compressed because the patient guards against probe pressure at that site; no discrete echogenic thrombus is seen, but complete compression also can't be demonstrated.

Naive read. A tech under time pressure documents "veins compress" for the full study, calls it negative for DVT, and moves to the next patient. The ordering physician discharges the patient believing DVT has been excluded.

Expert reasoning. Isolated distal (calf-vein) DVT accounts for a meaningful minority of lower-extremity DVT and roughly 15-25% of untreated calf-vein clots propagate proximally within about a week — which is exactly why the ACR-AIUM-SRU practice parameter for peripheral venous ultrasound calls for a repeat study in 5-7 days, rather than a same-day call of negative, when a segment is technically limited in a high pretest probability patient. A Wells score of 4 combined with one indeterminate calf segment does not meet the bar for calling the whole extremity negative — the study is negative in the segments actually assessed, and indeterminate at the one segment that matters most given the location of symptoms.

Deliverable — the worksheet impression, as written:

"FINDINGS: Common femoral, femoral, and popliteal veins are patent bilaterally with normal compressibility, respiratory phasicity, and augmentation. The right lower extremity venous system is normal throughout. On the left, the mid-calf posterior tibial venous segment could not be fully assessed for compressibility due to patient guarding at the site of measured swelling; no discrete thrombus is identified, but complete compression could not be demonstrated at this segment.

IMPRESSION: 1. No DVT identified in the visualized common femoral, femoral, or popliteal venous segments, bilaterally. 2. Technically limited assessment of the left mid-calf posterior tibial venous segment — indeterminate for distal DVT. Given a Wells score of 4 (high pretest probability) and this technical limitation, recommend D-dimer if not already obtained, or repeat duplex ultrasound of the left calf veins in 5-7 days if symptoms persist, per ACR-AIUM-SRU practice parameter guidance for indeterminate calf-vein assessment in high pretest probability patients."

Going deeper

• references/playbook.md — load when running a specific protocol: DVT compression sequence, RUQ/gallbladder major-minor criteria, graded-compression appendix technique, first-trimester dating/redating table, and organ-specific Doppler reference ranges.
• references/red-flags.md — load when a finding or artifact doesn't fit the expected pattern and you need the first question and the data to pull before escalating.
• references/vocabulary.md — load when a term of art needs to be used precisely, not approximately.

Sources

• SDMS/AIUM, Industry Standards for the Prevention of Work-Related Musculoskeletal Disorders in Sonography, 2017 revision (Journal of Diagnostic Medical Sonography, 33(5)) — source for the ~90% WRMSD prevalence figure and the ergonomic-technique guidance.
• AIUM, Practice Parameter for Documentation of an Ultrasound Examination, 2025 revision — source for the documentation and image-retention standard referenced throughout.
• AIUM/ACR/SRU, Practice Parameter for the Performance of Peripheral Venous Ultrasound Examination — source for the indeterminate calf-vein / repeat-in-5-7-days guidance in the worked example.
• Wells PS et al., Value of assessment of pretest probability of deep-vein thrombosis in clinical management, Lancet 1997, and the subsequent Wells criteria refinements — source for the DVT pretest-probability scoring used in the worked example.
• ACOG/AIUM/SMFM Committee Opinion No. 700, Methods for Estimating the Due Date — source for the CRL-based redating discrepancy thresholds in the playbook.
• Puylaert JB, Acute appendicitis: US evaluation using graded compression, Radiology 1986 — source for the graded-compression technique and the >6mm noncompressible-appendix criterion.
• Barakat MT et al. and subsequent AJR reviews on sonographic Murphy sign and gallbladder wall thickening — source for the major/minor cholecystitis criteria in the playbook.
• Rumack CM, Levine D (eds.), Diagnostic Ultrasound, 5th ed. (Elsevier) — standard reference textbook for protocol technique and organ-specific findings across subspecialties.
• ARDMS Scope of Practice and Clinical Preamble; SDMS Scope of Practice — source for the sonographer/physician division of diagnostic authority underlying the Identity and Communication style sections.
• Enrichment pass complete as of 2026; no direct practitioner sign-off on the role definition as a whole — flag via PR if you can confirm, correct, or add a citation.