Full FDA Non-Device CDS disclaimer

ARKA Clinical Decision Support is designed to meet all four criteria for Non-Device CDS under FD&C Act §520(o)(1)(E) and FDA's January 2026 final guidance on Clinical Decision Support Software. Recommendations support, not replace, the clinician's judgment. Every recommendation is anchored in a published guideline or peer-reviewed source, with the basis available for independent review. CLIN emphasizes imaging appropriateness at order entry.

This recommendation is intended to support, not replace, clinical judgment. It is generated by ARKA, software designed to meet the four criteria for Non-Device Clinical Decision Support under FD&C Act §520(o)(1)(E) and FDA's final guidance on Clinical Decision Support Software (January 2026). The clinician is responsible for the final decision.

Regulatory rationale Feature catalog Trust center

Champion toolkit

The 20-minute champion kit

Everything a physician champion needs to sponsor a VAC review — sponsor letter, colleague briefing, peer FAQ, and ARKA support commitments. Honest evidence posture only.

Public link for forwarding; not indexed for search. Pair with the VAC submission packet.

Block A

Why you

Denial pain relief today — and a real authorship path on the prospective study.

You feel the denial pain every day
Your orders bounce, rev-cycle chases documentation, and the rework lands on clinicians who already did the right study. ARKA targets the documentation gap at order entry — before the payer sees the claim — so the rework stops upstream.
Real academic credit on ARKA-PROSPECT-01
The prospective evaluation (ARKA-PROSPECT-01) is designed for co-authorship with site investigators — not a marketing poster. Protocol is drafted; IRB milestones update on the public evidence ladder. Sponsoring physicians who help shape the pilot cohort can earn authorship on the interim readout when TG-4 confirms the analysis plan.
Twenty minutes, not a writing project
ARKA pre-formats the VAC sponsor letter, committee deck outline, and peer FAQ below. You edit bracketed hospital fields, forward the kit to colleagues, and sign the attestation block — the committee gets honest evidence posture, not inflated pilot claims.

Block B

Pre-written from VAC packet sections B and D — fill bracketed hospital fields, sign the §I attestation block.

Physician champion letter

[Hospital Letterhead]
[Date]

To: Value Analysis Committee
From: [Champion Name], [Title], [Department/Service Line]
Re: Physician sponsorship — ARKA imaging CDS evaluation at [Hospital Name]

I am writing to sponsor ARKA Health's submission for committee review. At [Hospital Name], advanced-imaging prior authorization denials create rework for ordering clinicians and revenue-cycle staff — consistent with the [20–40%] industry band on high-cost imaging and our local baseline of [X%] on [MRI/CT/PET service line].

Clinical register — intended use (ARKA-CLIN, verbatim): ARKA-CLIN is intended for licensed health care professionals to support selection of clinically appropriate diagnostic imaging using structured clinical data. It presents ranked, evidence-cited imaging options drawn from published clinical guidelines and peer-reviewed literature, with the basis available for independent review. It does not acquire, process, or analyze medical images or physiological signals; does not provide time-critical alerts or triage; and does not place, cancel, or block orders.

Mechanism: Rules-first, guideline-anchored engine — every recommendation is anchored to a published guideline; if no guideline-anchored rule fires, ARKA-CLIN returns no card. Appropriateness comes from the transparent AIIE 2.0 core; an optional glass-box EBM supplies calibrated concordance probability only. The system falls back to the rule-based faithful reference when the calibrator is unavailable. Explicit negative scope: does not acquire, process, or analyze medical images or physiological signals; no time-critical alerts or triage; does not place, cancel, or block orders.

Current validation limitation, disclosed proactively: ARKA's published Tier-1 figure (~87.5% guideline concordance on a held-out signed-off cohort) is not a clinical-outcome claim. Synthetic self-consistency labels are plumbing-only and are never equated with clinical validity. Real-world retrospective evaluation is in progress at one site; prospective operational outcomes are the objective of ARKA-PROSPECT-01. We disclose this because committees should not discover limitations after deployment. Real-world retrospective evaluation results are public at /case-studies/stormont-vail-health; the prospective study design (ARKA-PROSPECT-01) is published on /evidence/methodology. Operational before/after outcomes require prospective use — we disclose this because committees should not discover limitations after deployment.

I have reviewed the VAC submission materials (/procurement/vac-packet) and sponsor their evaluation by this committee. A scoped [30–90 day] pilot in [one service line / one payer] is appropriate to test fit before the CY 2027 Promoting Interoperability electronic-PA attestation window.

---
Physician Champion attestation (VAC packet §I):
"I have reviewed this submission and sponsor its evaluation by the committee."

Name: ___________________________
Title / Department: ___________________________
Signature: ___________________________
Date: ___________________________

Block C

15-minute colleague briefing

Five slides: local denial pain → what ARKA does and doesn't → evidence ladder → pilot shape → the ask.

Briefing outline

1. Slide 1 — The denial problem locally (3 min)
- Open with your numbers: [X%] initial denial rate on [MRI/CT] for [top payer]; [Y] FTE-hours/month on rework (or one recent case your peers recognize).
- Industry context only — advanced imaging PA denials commonly run 20–40%; ~65% of denied claims are never reworked.
- CMS rescinded the mandatory AUC/PAMA program in 2024 — appropriateness is now enforced reactively through PA denials, not proactive CDS mandates.
- Ask: "What would we recover if even half of avoidable imaging denials left as clean first-pass claims?"
2. Slide 2 — What ARKA does / doesn't do (3 min)
- Does: scores appropriateness at order entry from structured FHIR; cites guidelines + SHAP reasoning; supports payer-side PA readiness (ARKA-INS layer).
- Does not: read image pixels; block, cancel, or modify orders; add clicks when no guideline fires; replace clinician judgment.
- Fail-open: if ARKA is unreachable, Epic proceeds exactly as today — no cards, no delays.
- Show the 30-second CDS Hooks demo or the one-page overview — not a live production claim.
3. Slide 3 — Evidence ladder, honestly (4 min)
- Tier 1 (published): ~87.5% guideline concordance on a held-out signed-off cohort — not a clinical-outcome claim. Synthetic self-consistency is plumbing only.
- Tier 2 (in progress): retrospective evaluation on 2,310 de-identified historical orders at a 586-bed Epic system — what ARKA would have flagged vs. what payers actually did.
- Tier 3 (planned): ARKA-PROSPECT-01 prospective silent-mode evaluation — the only tier that can claim operational before/after deltas.
- Point to /evidence/methodology and /docs/model-limitations — do not quote denial-rate "before/after" until Tier 3 exists.
4. Slide 4 — Pilot shape (3 min)
- 30–90 days, one service line or one payer — departmental OpEx, no capital request.
- Phases: paperwork → connect in sandbox → shadow mode with your sign-off → production enablement.
- Hospital lift: ~20–40 hrs EHR analyst, ~8–12 hrs security reviewer, ~4–8 hrs champion — not a dedicated project squad.
- Success KPIs co-signed up front; Pilot-to-Production guarantee if thresholds are missed.
5. Slide 5 — The ask (2 min)
- Ask for a 45-minute VAC prep session with radiology leadership + rev-cycle — ARKA attends and brings the pre-filled packet.
- Request a named rev-cycle contact for baseline denial extracts and a target VAC date [Month YYYY].
- Offer the champion kit link (/for/champions) and sponsor letter template for their signature block.
- Close: "We're not asking you to trust a 59% slide — we're asking for a scoped pilot with honest measurement."

Block D

Peer FAQ

Six questions colleagues ask before they sponsor you — answers from buyer FAQ and model-limitations copy.

Peer FAQ (copy all)

Will ARKA block or override my orders?: No. ARKA is non-blocking CDS — it cannot place, cancel, modify, or hold orders. If ARKA is unreachable, the EHR proceeds exactly as before: no cards render and no orders are delayed or blocked. Full deactivation is a configuration change (deregister CDS Hooks / SMART entries) with a documented rollback runbook. Learn more
Is this another alert-fatigue tool?: ARKA is silent unless a guideline-anchored rule fires — zero added workflow when nothing applies. Cards deliver in under 800ms in-flow; override and dismiss actions are logged. Alert burden is a first-class pilot KPI on /outcomes, reviewed weekly in shadow mode with your champion. Learn more
Who is liable if ARKA is wrong?: ARKA is Non-Device Clinical Decision Support under §520(o)(1)(E) of the FD&C Act. The ordering clinician retains full responsibility for the final decision — stated on every card. SHAP reasoning and cited guidelines support independent review (Criterion 4); ARKA is not for time-critical triage. Learn more
What evidence exists today — and what doesn't?: Tier 1: ~87.5% guideline concordance on a held-out signed-off cohort — not a clinical-outcome claim. Synthetic self-consistency labels smoke-test plumbing only. Tier 2: retrospective evaluation on real historical orders (in progress; public case study). Tier 3: prospective pilot planned — the only tier that can claim measured denial-rate or turnaround deltas. No ARKA page claims Tier 3 operational outcomes today. Learn more
How much of my time does this take?: A typical single-department pilot needs a clinical champion for ~4–8 hours total — sponsor letter, shadow-mode card review, and threshold sign-off. Ordering clinicians get a 15-minute orientation; champions/UM get a 45-minute session. ARKA drafts committee materials; you edit bracketed fields. Learn more
What happens to my feedback and overrides?: Accept, dismiss, and override actions are logged as first-class KPIs. Near-miss cards feed QI review on a weekly cadence during pilot shadow mode, then monthly with clinical governance. Override rates inform threshold tuning — your feedback shapes pilot configuration, not a black-box model retrain without release notes. Learn more

Block E

What ARKA does for you

You bring clinical credibility; we carry the committee prep.

We draft, you edit
VAC sponsor letter, committee talking points, security questionnaire seed answers, and the pre-filled submission packet — you fill bracketed hospital fields and clinical context, not a blank page.
We present to any committee
ARKA joins VAC, AI governance, security, and rev-cycle sessions — live or remote — with the honest evidence ladder, CMS-0057-F hospital brief, and model-limitations disclosure. You are not the sole technical explainer.
Named implementation engineer
First two weeks of production include daily check-ins and a named on-call engineer. Shadow mode and rollback runbook walkthrough happen before any clinician sees a card.
Reference-call protection
Named references are available to qualified prospects under mutual courtesy limits (≤2 calls/month per reference) via the access request form.

Reference policy details on References & evidence.

Loading…

Full FDA Non-Device CDS disclaimer

Regulatory rationale Feature catalog Trust center

Champion toolkit

The 20-minute champion kit

Everything a physician champion needs to sponsor a VAC review — sponsor letter, colleague briefing, peer FAQ, and ARKA support commitments. Honest evidence posture only.

Public link for forwarding; not indexed for search. Pair with the VAC submission packet.

Block A

Why you

Denial pain relief today — and a real authorship path on the prospective study.

You feel the denial pain every day
Your orders bounce, rev-cycle chases documentation, and the rework lands on clinicians who already did the right study. ARKA targets the documentation gap at order entry — before the payer sees the claim — so the rework stops upstream.
Real academic credit on ARKA-PROSPECT-01
The prospective evaluation (ARKA-PROSPECT-01) is designed for co-authorship with site investigators — not a marketing poster. Protocol is drafted; IRB milestones update on the public evidence ladder. Sponsoring physicians who help shape the pilot cohort can earn authorship on the interim readout when TG-4 confirms the analysis plan.
Twenty minutes, not a writing project
ARKA pre-formats the VAC sponsor letter, committee deck outline, and peer FAQ below. You edit bracketed hospital fields, forward the kit to colleagues, and sign the attestation block — the committee gets honest evidence posture, not inflated pilot claims.

Block B

Pre-written from VAC packet sections B and D — fill bracketed hospital fields, sign the §I attestation block.

Physician champion letter

[Hospital Letterhead]
[Date]

To: Value Analysis Committee
From: [Champion Name], [Title], [Department/Service Line]
Re: Physician sponsorship — ARKA imaging CDS evaluation at [Hospital Name]

I am writing to sponsor ARKA Health's submission for committee review. At [Hospital Name], advanced-imaging prior authorization denials create rework for ordering clinicians and revenue-cycle staff — consistent with the [20–40%] industry band on high-cost imaging and our local baseline of [X%] on [MRI/CT/PET service line].

Clinical register — intended use (ARKA-CLIN, verbatim): ARKA-CLIN is intended for licensed health care professionals to support selection of clinically appropriate diagnostic imaging using structured clinical data. It presents ranked, evidence-cited imaging options drawn from published clinical guidelines and peer-reviewed literature, with the basis available for independent review. It does not acquire, process, or analyze medical images or physiological signals; does not provide time-critical alerts or triage; and does not place, cancel, or block orders.

Mechanism: Rules-first, guideline-anchored engine — every recommendation is anchored to a published guideline; if no guideline-anchored rule fires, ARKA-CLIN returns no card. Appropriateness comes from the transparent AIIE 2.0 core; an optional glass-box EBM supplies calibrated concordance probability only. The system falls back to the rule-based faithful reference when the calibrator is unavailable. Explicit negative scope: does not acquire, process, or analyze medical images or physiological signals; no time-critical alerts or triage; does not place, cancel, or block orders.

Current validation limitation, disclosed proactively: ARKA's published Tier-1 figure (~87.5% guideline concordance on a held-out signed-off cohort) is not a clinical-outcome claim. Synthetic self-consistency labels are plumbing-only and are never equated with clinical validity. Real-world retrospective evaluation is in progress at one site; prospective operational outcomes are the objective of ARKA-PROSPECT-01. We disclose this because committees should not discover limitations after deployment. Real-world retrospective evaluation results are public at /case-studies/stormont-vail-health; the prospective study design (ARKA-PROSPECT-01) is published on /evidence/methodology. Operational before/after outcomes require prospective use — we disclose this because committees should not discover limitations after deployment.

I have reviewed the VAC submission materials (/procurement/vac-packet) and sponsor their evaluation by this committee. A scoped [30–90 day] pilot in [one service line / one payer] is appropriate to test fit before the CY 2027 Promoting Interoperability electronic-PA attestation window.

---
Physician Champion attestation (VAC packet §I):
"I have reviewed this submission and sponsor its evaluation by the committee."

Name: ___________________________
Title / Department: ___________________________
Signature: ___________________________
Date: ___________________________

Block C

15-minute colleague briefing

Five slides: local denial pain → what ARKA does and doesn't → evidence ladder → pilot shape → the ask.

Briefing outline

1. Slide 1 — The denial problem locally (3 min)
- Open with your numbers: [X%] initial denial rate on [MRI/CT] for [top payer]; [Y] FTE-hours/month on rework (or one recent case your peers recognize).
- Industry context only — advanced imaging PA denials commonly run 20–40%; ~65% of denied claims are never reworked.
- CMS rescinded the mandatory AUC/PAMA program in 2024 — appropriateness is now enforced reactively through PA denials, not proactive CDS mandates.
- Ask: "What would we recover if even half of avoidable imaging denials left as clean first-pass claims?"
2. Slide 2 — What ARKA does / doesn't do (3 min)
- Does: scores appropriateness at order entry from structured FHIR; cites guidelines + SHAP reasoning; supports payer-side PA readiness (ARKA-INS layer).
- Does not: read image pixels; block, cancel, or modify orders; add clicks when no guideline fires; replace clinician judgment.
- Fail-open: if ARKA is unreachable, Epic proceeds exactly as today — no cards, no delays.
- Show the 30-second CDS Hooks demo or the one-page overview — not a live production claim.
3. Slide 3 — Evidence ladder, honestly (4 min)
- Tier 1 (published): ~87.5% guideline concordance on a held-out signed-off cohort — not a clinical-outcome claim. Synthetic self-consistency is plumbing only.
- Tier 2 (in progress): retrospective evaluation on 2,310 de-identified historical orders at a 586-bed Epic system — what ARKA would have flagged vs. what payers actually did.
- Tier 3 (planned): ARKA-PROSPECT-01 prospective silent-mode evaluation — the only tier that can claim operational before/after deltas.
- Point to /evidence/methodology and /docs/model-limitations — do not quote denial-rate "before/after" until Tier 3 exists.
4. Slide 4 — Pilot shape (3 min)
- 30–90 days, one service line or one payer — departmental OpEx, no capital request.
- Phases: paperwork → connect in sandbox → shadow mode with your sign-off → production enablement.
- Hospital lift: ~20–40 hrs EHR analyst, ~8–12 hrs security reviewer, ~4–8 hrs champion — not a dedicated project squad.
- Success KPIs co-signed up front; Pilot-to-Production guarantee if thresholds are missed.
5. Slide 5 — The ask (2 min)
- Ask for a 45-minute VAC prep session with radiology leadership + rev-cycle — ARKA attends and brings the pre-filled packet.
- Request a named rev-cycle contact for baseline denial extracts and a target VAC date [Month YYYY].
- Offer the champion kit link (/for/champions) and sponsor letter template for their signature block.
- Close: "We're not asking you to trust a 59% slide — we're asking for a scoped pilot with honest measurement."

Block D

Peer FAQ

Six questions colleagues ask before they sponsor you — answers from buyer FAQ and model-limitations copy.

Peer FAQ (copy all)

Will ARKA block or override my orders?: No. ARKA is non-blocking CDS — it cannot place, cancel, modify, or hold orders. If ARKA is unreachable, the EHR proceeds exactly as before: no cards render and no orders are delayed or blocked. Full deactivation is a configuration change (deregister CDS Hooks / SMART entries) with a documented rollback runbook. Learn more
Is this another alert-fatigue tool?: ARKA is silent unless a guideline-anchored rule fires — zero added workflow when nothing applies. Cards deliver in under 800ms in-flow; override and dismiss actions are logged. Alert burden is a first-class pilot KPI on /outcomes, reviewed weekly in shadow mode with your champion. Learn more
Who is liable if ARKA is wrong?: ARKA is Non-Device Clinical Decision Support under §520(o)(1)(E) of the FD&C Act. The ordering clinician retains full responsibility for the final decision — stated on every card. SHAP reasoning and cited guidelines support independent review (Criterion 4); ARKA is not for time-critical triage. Learn more
What evidence exists today — and what doesn't?: Tier 1: ~87.5% guideline concordance on a held-out signed-off cohort — not a clinical-outcome claim. Synthetic self-consistency labels smoke-test plumbing only. Tier 2: retrospective evaluation on real historical orders (in progress; public case study). Tier 3: prospective pilot planned — the only tier that can claim measured denial-rate or turnaround deltas. No ARKA page claims Tier 3 operational outcomes today. Learn more
How much of my time does this take?: A typical single-department pilot needs a clinical champion for ~4–8 hours total — sponsor letter, shadow-mode card review, and threshold sign-off. Ordering clinicians get a 15-minute orientation; champions/UM get a 45-minute session. ARKA drafts committee materials; you edit bracketed fields. Learn more
What happens to my feedback and overrides?: Accept, dismiss, and override actions are logged as first-class KPIs. Near-miss cards feed QI review on a weekly cadence during pilot shadow mode, then monthly with clinical governance. Override rates inform threshold tuning — your feedback shapes pilot configuration, not a black-box model retrain without release notes. Learn more

Block E

What ARKA does for you

You bring clinical credibility; we carry the committee prep.

We draft, you edit
VAC sponsor letter, committee talking points, security questionnaire seed answers, and the pre-filled submission packet — you fill bracketed hospital fields and clinical context, not a blank page.
We present to any committee
ARKA joins VAC, AI governance, security, and rev-cycle sessions — live or remote — with the honest evidence ladder, CMS-0057-F hospital brief, and model-limitations disclosure. You are not the sole technical explainer.
Named implementation engineer
First two weeks of production include daily check-ins and a named on-call engineer. Shadow mode and rollback runbook walkthrough happen before any clinician sees a card.
Reference-call protection
Named references are available to qualified prospects under mutual courtesy limits (≤2 calls/month per reference) via the access request form.

Reference policy details on References & evidence.

The 20-minute champion kit

Why you

You feel the denial pain every day

Real academic credit on ARKA-PROSPECT-01

Twenty minutes, not a writing project

VAC sponsor letter template

Physician champion letter

15-minute colleague briefing

Briefing outline

Peer FAQ

Peer FAQ (copy all)

What ARKA does for you

We draft, you edit

We present to any committee

Named implementation engineer

Reference-call protection

The 20-minute champion kit

Why you

You feel the denial pain every day

Real academic credit on ARKA-PROSPECT-01

Twenty minutes, not a writing project

VAC sponsor letter template

Physician champion letter

15-minute colleague briefing

Briefing outline

Peer FAQ

Peer FAQ (copy all)

What ARKA does for you

We draft, you edit

We present to any committee

Named implementation engineer

Reference-call protection