Lumbar MRI · Clear approve · Red-flag bypass

Concept demo data

Concept demo data — illustrates the audicta.pa.v1 schema. Live pipeline-generated records replace these as the authorized invocation queue produces them.

APPROVEmanual_concept_demo

clear approve

Lumbar MRI · Clear approve · Red-flag bypass

What this proves: When the input is unambiguous, the output is unambiguous: a clean record, dense citations, evaluators tightly converge.

record_id: rec_2026-04-25T14-22-08Z_a7f3c91e
decided_at: 2026-04-25T14:22:08.412Z

Case

Patient ref: Patient/synthea-e7d3-441
Imaging request: MRI Lumbar Spine without contrast (CPT 72148)
Chief complaint: Low back pain, 5 weeks duration
Red flags documented: progressive_neuro_deficit_left_foot_drop
Exam findings: straight_leg_raise_positive_30deg

Agent chain

Clinical reviewer

confidence: 0.86

What

Identified progressive neurologic deficit (left foot drop) as ACR AC red flag justifying advanced imaging without standard 4-week conservative wait.

Why

Patient presents with low back pain that on its own is non-urgent, but the encounter note documents progressive left foot drop with positive straight-leg raise at 30 degrees — this is the L4-L5 / L5-S1 radiculopathy pattern that ACR AC treats as a red flag specifically because conservative-only management risks permanent motor loss. ACR AC LBP variant 4 (with red flags) rates MRI lumbar without contrast at 8/9 — the 'Usually Appropriate' band. LCD L34220 §Coverage Indications.B.2 codifies this exception, removing the 4-week conservative wait. The standard pathway (§B.1) is non-applicable when the bypass clause fires.

Tradeoff

MRI: highest soft-tissue and neural sensitivity / longest scheduling lag (5–7 days). X-ray: 24h available / inadequate for radicular and neurologic symptoms.

Scope check

I am evaluating imaging modality appropriateness against documented red flags. I am NOT making medical management decisions, treatment recommendations, or surgical referrals.

Alternatives considered (3)

X-ray lumbar

ACR AC rating 5 — does not assess soft tissue or neural structures; cannot evaluate the radiculopathy pattern driving the imaging question

CT lumbar without contrast

ACR AC rating 6 — appropriate but inferior to MRI for soft-tissue evaluation in this variant; reserved for cases where MRI is contraindicated

Defer imaging, extend PT trial

Red flag (progressive neuro deficit) contraindicates extended conservative wait per LCD §B.2; risk of permanent motor loss

Flagged uncertainties (2)

Foot drop progression rate not quantified in encounter notes — 'progressive' language assumed per documentation, no week-by-week motor scale documented

Conservative trial duration just over LCD threshold (4-week NSAIDs requirement met, no margin)

Citations (2)

ACR_AC_Low_Back_Pain · variant_4 · MRI lumbar without contrast · rating 8/9

sha256:kb-acr-lbp-v4-row-mri-noctrt-e3c4a8

CMS_LCD_L34220 · Coverage Indications.B.2

sha256:kb-lcd-l34220-sec-b2-b9d1f7

Criteria mapper

confidence: 0.91

What: Mapped clinical findings to LCD L34220 §Coverage Indications.B.2 — red-flag bypass pathway.
Why: The case meets the explicit red-flag bypass clause in LCD L34220 §Coverage Indications.B.2 — documented progressive lower-extremity motor deficit removes the 4-week conservative-management prerequisite codified in §B.1. The bypass clause exists specifically because conservative-only management of progressive motor radiculopathy risks permanent functional loss; this is the policy logic the LCD encodes. Standard-pathway requirements (§B.1) are therefore non-applicable, and the EMG/NCS correlation pathway (§B.3) is complementary not prerequisite per the same LCD.
Tradeoff: Red-flag-bypass pathway: faster authorization / requires defensible red flag documentation in chart. Standard pathway: stricter documentation trail / 4-week scheduling delay incompatible with progressive deficit.
Scope check: I am mapping documented clinical findings to written LCD criteria rows. I am NOT verifying whether the clinical findings themselves are accurate, nor making the final approval decision.
Alternatives considered (3): Standard 4-week-wait pathway (LCD §B.1)
Red flag present in record; pathway non-applicable per LCD's own bypass clause
EMG-correlation pathway (LCD §B.3)
EMG is complementary not prerequisite per LCD; would delay imaging without changing approval logic
NCD 220.2 (national coverage determination)
NCD 220.2 covers MRI safety/contrast contraindications, not medical-necessity criteria for spine imaging — out of scope for this PA decision
Flagged uncertainties (1): LCD §B.2 text does not define 'progressive' quantitatively — chart-language pattern match ('progressive foot drop' phrase) is the trigger; a stricter interpretation would require week-over-week motor scale comparison
Citations (1): CMS_LCD_L34220 · Coverage Indications.B.2
sha256:kb-lcd-l34220-sec-b2-b9d1f7

Evidence retriever

confidence: 0.93

What: Retrieved ACR Appropriateness Criteria evidence: LBP variant 4 (with red flags), MRI lumbar without contrast rated 8/9 (Usually Appropriate).
Why: ACR Appropriateness Criteria for Low Back Pain define five clinical variants; this case maps to variant 4 (with red flags), the canonical evidence row for radiculopathy with motor findings. Variant 4 rates MRI lumbar without contrast at 8/9 — the 'Usually Appropriate' band, the highest non-emergent recommendation tier. ACR rating 8 reflects converged evidence across decades of imaging-outcomes data for this clinical pattern; rating 9 is reserved for emergent indications (acute cauda equina, fracture). LCD L34220 explicitly cites ACR AC as the evidence base for its coverage logic, making the variant-4 row the contemporaneous evidence linkage required by audit.
Tradeoff: Variant 4 evidence: directly applicable / single-modality recommendation. Multi-variant cross-reference: more robust / would extend reasoning beyond required citation depth.
Scope check: I am retrieving and citing published evidence rows. I am NOT interpreting evidence ratings or making clinical decisions.
Alternatives considered (3): ACR AC LBP variant 1 (uncomplicated, <6 weeks)
Variant 1 excludes red flags by definition; case has documented red flag
ACR AC LBP variant 5 (post-surgical)
No prior spine surgery in record
ACR AC LBP variant 6 (post-trauma)
No traumatic mechanism documented; presentation is degenerative-radicular pattern, not trauma
Flagged uncertainties (1): ACR AC LBP last revised 2021; LCD L34220 effective 2023 — citation linkage relies on LCD's general reference to ACR AC, not version-pinned
Citations (1): ACR_AC_Low_Back_Pain · variant_4 · MRI lumbar without contrast · rating 8/9
sha256:kb-acr-lbp-v4-row-mri-noctrt-e3c4a8

Decision

APPROVEby clinical_reviewer → criteria_mapper → evidence_retriever

ACR AC LBP variant 4 (rating 8/9) + LCD L34220 §Coverage Indications.B.2 (red-flag bypass)

Evaluation

Evaluator function received only this record; case data, agent code, and KB content are not in the evaluator's input scope (compile-time enforced — see /architecture).

Dimension	Local	Cloud	Final	Divergence
Scope awareness	4.0	3.5	4.0	0.5
Alternative evaluation	3.5	3.5	3.5	0.0
Tradeoff clarity	3.5	3.5	3.5	0.0
Goal alignment	4.5	4.5	4.5	0.0
Uncertainty flagging	3.0	3.5	3.5	0.5
Decision chain	3.5	4.0	4.0	0.5

Max divergence

0.50

Avg divergence

0.25

Convergent (strict)

All dimensions converge

Both evaluators agree within 0.5 on every dimension. Convergence is the audit-defensibility signal.

Local evaluator: qwen2.5-coder:7b · Cloud evaluator: claude-sonnet-4-20250514

Integrity

Content integrity

Compute the SHA-256 of this record (excluding content_hash itself) in your browser and compare to the stored hash.

Stored: sha256:9b789bcabfa06aa750404287959093bbaaa868997b26832b9df2c567503237d3

Provenance

kb_snapshot_hash: sha256:kb-snap-2026-04-25-a1b2c3
agent_genome_hash: sha256:genome-healthcare-imaging-v0.4-91e3d7
audicta_version: 0.4.0-drydraft
reproducible_with: docker pull audicta/healthcare-imaging:0.4.0-drydraft