Source Register

Sources ARKA uses — what each costs to use, and who paid for it to exist.

Last updated: August 14, 2026

Overview

Vol III §25.1 item 12 — one page for the external sources that inform a shipped number or clinical surface, with type, licence status, conflict-of-interest class, GRADE certainty where it applies, evidence domain (imaging vs analogous-domain), version, and date. This is where the ACR-labels question, the ACR/Optum error, and the guidelines question are answered at once: if it is not here, it does not influence a shipped ARKA figure.

Related: How ARKA validates · /risk · /baseline · Evaluation criteria.

72 classified sources. Canonical machine artefact: lib/sources/data.json.

Single-specialty vs primary evidence

Where a single-specialty source and primary evidence both assign a 1–9 rating to the same cell, they diverge in 0 of 0 paired cells. Single-specialty sources rate imaging higher than primary evidence in 0 of those cells, and lower in 0. Mean absolute gap among diverging cells is 0 points. This is a finding about the field, not a claim that one source type is correct.

Single-specialty versus primary-evidence rating divergence
Slice	Paired cells (n)	Diverge	Higher than primary	Lower than primary	Mean \|Δ\|
Single-specialty vs primary	0	0	0	0	0

Where a single-specialty source and primary evidence both assign a 1–9 rating to the same cell, a gap of two or more points is a named divergence. The primary rating governs; the down-weight that resolved the disagreement is shown. Cells without a paired per-source rating cannot enter the table. This is a finding about the field, not a claim that one source type is correct. Method.

Register

By licence status

public domain30
open licence4
licensed to ARKA0
cited only, not redistributed35
not licensed — referenced by citation3

By COI class

none34
society with a financial interest in the service7
vendor-authored3
payer-authored0
government28

By GRADE certainty

high0
moderate6
low14
very low2
not rated (N/A)50

By evidence domain

imaging8
analogous-domain12
methodological1

We rate certainty only where certainty is a meaningful concept. A dataset is not evidence for a claim; it is the raw material for one.

SearchTypeLicenceCOIGRADEDomain

Showing 72 of 72 classified sources.

guideline (1)

Source	Domain	Licence	COI	GRADE	Version	Date	Used by
USPSTF D-rated services USPSTF Legacy licence text: `federal-public` GRADE — moderate: USPSTF D ratings require at least moderate certainty of no net benefit or net harm for the rated service; ARKA cites the D-rated service catalogue for deferred expansion, not a pooled nudge effect.	—	public domain	government	GRADE moderate USPSTF D ratings require at least moderate certainty of no net benefit or net harm for the rated service; ARKA cites the D-rated service catalogue for deferred expansion, not a pooled nudge effect.	USPSTF current as of retrieval	2026-08-07	ARKA-TCOC lvs-cervical-over-65 lvs-psa-over-75

measure specification (5)

Source	Domain	Licence	COI	GRADE	Version	Date	Used by
Change Healthcare 2020 Denials Index Change Healthcare / Rivet Health (secondary summary) Vendor industry denials index; cited for avoidable-denial rates, not used as a clinical measure codeset. Not applicable — this is a measure specification or codeset, not an effect estimate.	—	cited only, not redistributed	vendor-authored	—	as cited in lib/roi/roi-model.ts	2026-08-11	ROI
Milliman MedInsight Health Waste Calculator (with VBID Health) Milliman / VBID Health Legacy licence text: `licensed-do-not-embed` — do not embed, reverse-engineer or “model” Not applicable — this is a measure specification or codeset, not an effect estimate.	—	not licensed — referenced by citation	vendor-authored	—	Named comparator only; no version embedded	2026-08-07	ARKA-TCOC
NCQA HEDIS — e.g. Use of Imaging Studies for Low Back Pain (LBP) NCQA Professional society that licenses the named code set or measure product; ARKA cites identifiers only and does not redistribute the corpus. Legacy licence text: `licensed-do-not-embed` — reference by name only; do not ship value sets Not applicable — this is a measure specification or codeset, not an effect estimate.	—	not licensed — referenced by citation	society with a financial interest in the service	—	HEDIS MY as published by NCQA (not embedded)	2026-08-07	ARKA-TCOC
Schwartz AL, Landon BE, Elshaug AG, Chernew ME, McWilliams JM. Measuring Low-Value Care in Medicare. JAMA Intern Med. 2014;174(7):1067-76. CMS / HHS Professional society that licenses the named code set or measure product; ARKA cites identifiers only and does not redistribute the corpus. Legacy licence text: `public-academic` Primary anchor for measure definitions: lvi-asympt-cardiac, lvi-carotid-asympt, lvi-headache-imaging, lvi-lbp-imaging, lvi-msk-early-mri, lvi-preop-cxr, lvi-preop-echo, lvi-preop-stress, lvi-repeat-ctap, lvi-sinusitis-imaging, lvi-syncope-imaging. Not applicable — this is a measure specification or codeset, not an effect estimate.	—	cited only, not redistributed	society with a financial interest in the service	—	2014 print; methods supplement	2026-08-07	ARKA-TCOC lvi-asympt-cardiac lvi-carotid-asympt lvi-headache-imaging lvi-lbp-imaging lvi-msk-early-mri lvi-preop-cxr lvi-preop-echo lvi-preop-stress lvi-repeat-ctap lvi-sinusitis-imaging lvi-syncope-imaging
CollaboRATE: a fast and frugal patient-reported measure of shared decision making The Dartmouth Institute for Health Policy & Clinical Practice (Elwyn / Barr) Registered for use with the CollaboRATE authors (registration form). Three-item patient-reported shared-decision measure; ARKA adopts the published top-score (top-box) rule without modifying item wording, response scale, or scoring. Not applicable — this is a measure specification or instrument, not an effect estimate.	—	open licence	none	—	CollaboRATE adult 0–9 top-score scoring (published)	2013-01-01	ARKA-SDM lib/sdm/measurement.ts lib/appropriateness/dimensions.ts

peer-reviewed trial (21)

Source	Domain	Licence	COI	GRADE	Version	Date	Used by
Bhayana et al., Radiology 2025 (314:2) — "Leveraging LLMs to Generate Clinical Histories for Oncologic Imaging Requisitions" Radiology (RSNA) Independent peer-reviewed academic publication (or Cochrane review); not society-, vendor-, or payer-authored. Legacy licence text: cited only GRADE — very low: observational LLM evaluation (clinical-history generation for oncologic imaging requisitions); non-randomised, single-study, technology-specific — prior art for IRE, not an ARKA intervention effect.	imaging	cited only, not redistributed	none	GRADE very low observational LLM evaluation (clinical-history generation for oncologic imaging requisitions); non-randomised, single-study, technology-specific — prior art for IRE, not an ARKA intervention effect.	as published	2026-08-08	ARKA-IRE
Brehaut JC et al. Practice Feedback Interventions: 15 Suggestions for Optimizing Effectiveness. Ann Intern Med. 2016;164(6):435-441. Annals of Internal Medicine Independent peer-reviewed academic publication (or Cochrane review); not society-, vendor-, or payer-authored. Legacy licence text: Published academic Practice-feedback design guidance (15 suggestions); cited for peer-comparison packet design. GRADE — low: non-randomised design-guidance synthesis (15 suggestions); starts low as expert/methods consensus drawn from A&F experience, not a trial of those suggestions as an intervention package.	analogous-domain	cited only, not redistributed	none	GRADE low non-randomised design-guidance synthesis (15 suggestions); starts low as expert/methods consensus drawn from A&F experience, not a trial of those suggestions as an intervention package.	2016;164(6)	2026-08-07	/evidence/nudges /risk ARKA-TCOC
Chung Y, Nicola LP, Nicola GN, Rula EY. Imaging Utilization Differences After Telemedicine Versus In-Person Visits. J Am Coll Radiol. 2025 Jul;22(7):1047-1055. Journal of the American College of Radiology Harvey L. Neiman Health Policy Institute / Radiology Associates authors. Optum Clinformatics matched case-control. 7-day post-visit imaging 2.4 pp lower after telemedicine (29.7% relative reduction). JACR is an American College of Radiology journal — COI class society with financial interest in the service. Volume/utilization outcome, not appropriateness rating. GRADE — low: peer-reviewed publication typed as a trial/evidence source; defaulted to low pending indication-specific grading (not rated high).	imaging	cited only, not redistributed	society with a financial interest in the service	GRADE low peer-reviewed publication typed as a trial/evidence source; defaulted to low pending indication-specific grading (not rated high).	2025 Jul;22(7):1047-1055; 23,431,032 matched office visits (2021)	2025-07-17	/evidence/methodology docs/governance/VIRTUAL_ENCOUNTER_POSITION.md
Clark-Randall A, Halpern DJ, Taylor J, Roth CJ, Gupta RT, Woodall J, Shah KP. Implementation of a Radiology Utilization Dashboard Yields Significant Cost Savings in a Large Primary Care Network. J Am Coll Radiol. 2021 Jul;18(7):947–950. Journal of the American College of Radiology Independent published source cited on /evidence/nudges and /risk; not redistributed. Independent academic authors (Duke University / Duke Primary Care / Duke Radiology). PubMed record has no abstract and no GrantList. No vendor funding disclosed in the bibliographic record. COI class assigned from author affiliations and the PubMed record — full text not read. Same Duke Primary Care cohort. One result reported on two denominators — volume (halpern-2021) and cost (clark-randall-2021). Not two independent findings. Overlap is inferred from shared network, overlapping authors (Halpern, Clark-Randall, Woodall, Shah), and the cost paper citing the volume paper; the JACR full text was not read. GRADE — low: observational pre-post design, no concurrent control group, no reported risk adjustment, single health system, effect not separable from secular trend. Effect size is not a reason to upgrade — the size of an effect and the certainty of an effect are different quantities.	imaging	cited only, not redistributed	none	GRADE low observational pre-post design, no concurrent control group, no reported risk adjustment, single health system, effect not separable from secular trend. Effect size is not a reason to upgrade — the size of an effect and the certainty of an effect are different quantities.	2021 Jul;18(7):947–950; cost denominator of the Duke Primary Care dashboard	2021-07-01	/evidence/nudges /risk ARKA-TCOC
Eur Radiol 2025 — "Applying large language model for automated quality scoring of radiology requisitions using standardized criteria" Radiology (RSNA) Independent peer-reviewed academic publication (or Cochrane review); not society-, vendor-, or payer-authored. Legacy licence text: cited only GRADE — very low: observational LLM evaluation (automated requisition quality scoring); non-randomised, single-study — prior art for IRE, not an ARKA intervention effect.	imaging	cited only, not redistributed	none	GRADE very low observational LLM evaluation (automated requisition quality scoring); non-randomised, single-study — prior art for IRE, not an ARKA intervention effect.	as published	2026-08-08	ARKA-IRE
Halpern DJ, Clark-Randall A, Woodall J, Anderson J, Shah K. Reducing Imaging Utilization in Primary Care Through Implementation of a Peer Comparison Dashboard. J Gen Intern Med. 2021 Jan;36(1):108–113. Journal of General Internal Medicine Independent published source cited on /evidence/nudges and /risk; not redistributed. Independent academic authors (Duke University / Duke Primary Care). No vendor funding disclosed in the abstract or in the available full-text extract (PMC7859117). PubMed record has no GrantList. COI class assigned from the abstract plus the available full-text extract; a dedicated funding statement was not located in the extract. Same Duke Primary Care cohort. One result reported on two denominators — volume (halpern-2021) and cost (clark-randall-2021). Not two independent findings. Overlap is inferred from shared network, overlapping authors (Halpern, Clark-Randall, Woodall, Shah), and the cost paper citing the volume paper; the JACR full text was not read. GRADE — low: observational pre-post design, no concurrent control group, no reported risk adjustment, single health system, effect not separable from secular trend. Effect size is not a reason to upgrade — the size of an effect and the certainty of an effect are different quantities.	imaging	cited only, not redistributed	none	GRADE low observational pre-post design, no concurrent control group, no reported risk adjustment, single health system, effect not separable from secular trend. Effect size is not a reason to upgrade — the size of an effect and the certainty of an effect are different quantities.	2021; 159 providers / 1,358,644 outpatient encounters / 3 years	2021-01-01	/evidence/nudges /risk ARKA-TCOC
Ivers N et al. Audit and feedback: effects on professional practice (Cochrane). Cochrane Collaboration Independent peer-reviewed academic publication (or Cochrane review); not society-, vendor-, or payer-authored. Legacy licence text: Published / Cochrane Cochrane systematic review of audit-and-feedback effects on professional practice. GRADE — moderate: Cochrane systematic review of randomised audit-and-feedback trials; starts high, downgraded one level for inconsistency/heterogeneity across interventions and settings (and residual indirectness to imaging utilisation).	analogous-domain	cited only, not redistributed	none	GRADE moderate Cochrane systematic review of randomised audit-and-feedback trials; starts high, downgraded one level for inconsistency/heterogeneity across interventions and settings (and residual indirectness to imaging utilisation).	Most recently updated 2025 per build book	2026-08-07	/evidence/nudges /risk ARKA-TCOC
Johns Hopkins — commercial vs. Medicare radiology prices (2021) Johns Hopkins University Academic price comparison study; independent of society/vendor/payer authorship of ARKA measures. GRADE — low: observational price-comparison study (commercial vs Medicare radiology); starts low as non-randomised evidence; used for ROI unit prices, not as an intervention effect.	imaging	cited only, not redistributed	none	GRADE low observational price-comparison study (commercial vs Medicare radiology); starts low as non-randomised evidence; used for ROI unit prices, not as an intervention effect.	as cited in lib/roi/roi-model.ts	2026-08-11	ROI
Johnson EJ, Goldstein D. Medicine. Do defaults save lives? Science. 2003;302(5649):1338–1339. Science Defaults / organ-donation framing study; cited for default-effect transfer, not imaging-specific effect. GRADE — low: observational / framing evidence on defaults (organ donation); starts low as non-randomised, high indirectness to imaging order defaults.	analogous-domain	cited only, not redistributed	none	GRADE low observational / framing evidence on defaults (organ donation); starts low as non-randomised, high indirectness to imaging order defaults.	as cited in lib/tcoc/content/nudge-evidence.ts	2026-08-12	ARKA-TCOC /evidence/nudges /risk
Liu T, Zhu Z, Thompson MP, McCullough JS, Hou H, Chang C-H, Fendrick AM, Ellimoottil C. Primary Care Practice Telehealth Use and Low-Value Care Services. JAMA Netw Open. 2024 Nov 7;7(11):e2445436. JAMA Network Open Medicare FFS cohort (2019–2022). High practice-level telehealth use was associated with lower rates of some office-based low-value services and no association with higher imaging-based low-value measures. Independent academic authors (University of Michigan). COI class from abstract and affiliations — full conflict statement not independently verified. GRADE — low: peer-reviewed publication typed as a trial/evidence source; defaulted to low pending indication-specific grading (not rated high).	analogous-domain	cited only, not redistributed	none	GRADE low peer-reviewed publication typed as a trial/evidence source; defaulted to low pending indication-specific grading (not rated high).	2024 Nov 7;7(11):e2445436; 577,928 beneficiaries / 2,552 Michigan primary care practices	2024-11-07	/evidence/methodology docs/governance/VIRTUAL_ENCOUNTER_POSITION.md
Meeker D et al. durability follow-up. JAMA. 2017. JAMA Network Independent peer-reviewed academic publication (or Cochrane review); not society-, vendor-, or payer-authored. Legacy licence text: Published academic Independent published source cited on /evidence/nudges and /risk; not redistributed. GRADE — low: prespecified secondary analysis of the Meeker 2016 cluster-RCT (Linder et al. 2017), 12-month post-removal follow-up only; downgraded for indirectness (antibiotics → imaging) and imprecision (single follow-up window, decaying effects).	analogous-domain	cited only, not redistributed	none	GRADE low prespecified secondary analysis of the Meeker 2016 cluster-RCT (Linder et al. 2017), 12-month post-removal follow-up only; downgraded for indirectness (antibiotics → imaging) and imprecision (single follow-up window, decaying effects).	2017 (12 months post-removal)	2026-08-07	/evidence/nudges /risk ARKA-TCOC
Meeker D et al. Effect of Behavioral Interventions on Inappropriate Antibiotic Prescribing. JAMA. 2016. JAMA Network Independent peer-reviewed academic publication (or Cochrane review); not society-, vendor-, or payer-authored. Legacy licence text: Published academic Independent cluster-RCT (Meeker et al. 2016); also listed from the TCOC register. Independent published source cited on /evidence/nudges and /risk; not redistributed. GRADE — moderate: randomised, low risk of bias; downgraded one level for indirectness — the population and the clinical decision are antibiotics, not imaging.	analogous-domain	cited only, not redistributed	none	GRADE moderate randomised, low risk of bias; downgraded one level for indirectness — the population and the clinical decision are antibiotics, not imaging.	2016; 47 practices / 248 clinicians / 18 months	2026-08-07	/evidence/nudges /risk ARKA-TCOC ROI
Meeker D et al. JAMA 2016 (accountable justification arm) American Medical Association Professional society that licenses the named code set or measure product; ARKA cites identifiers only and does not redistribute the corpus. Legacy licence text: Published academic GRADE — moderate: cluster-randomised, low risk of bias; downgraded one level for indirectness — accountable justification was tested on antibiotic prescribing, not imaging orders.	analogous-domain	cited only, not redistributed	society with a financial interest in the service	GRADE moderate cluster-randomised, low risk of bias; downgraded one level for indirectness — accountable justification was tested on antibiotic prescribing, not imaging orders.	2016	2026-08-09	ARKA-AJ
Meeker D, Knight TK, Friedberg MW, et al. Nudging Guideline-Concordant Antibiotic Prescribing: A Randomized Clinical Trial. JAMA Intern Med. 2014;174(3):425–431. JAMA Internal Medicine Cluster-RCT of nudging guideline-concordant antibiotic prescribing (Meeker et al. 2014); precursor to the 2016 three-arm trial. GRADE — moderate: cluster-randomised antibiotic-prescribing nudge trial; downgraded one level for indirectness to imaging utilisation.	analogous-domain	cited only, not redistributed	none	GRADE moderate cluster-randomised antibiotic-prescribing nudge trial; downgraded one level for indirectness to imaging utilisation.	as cited in lib/tcoc/content/nudge-evidence.ts	2026-08-12	ARKA-TCOC /evidence/nudges /risk
Patel MS, Day SC, Halpern SD, et al. Generic Medication Prescription Rates After Health System-Wide Redesign of Default Options Within the Electronic Health Record. JAMA Intern Med. 2016;176(6):847–848. JAMA Internal Medicine Health-system EHR default redesign study (generic prescribing); cited for default-effect mechanism. GRADE — low: health-system EHR default redesign study (before/after); non-randomised, different clinical decision (generics), cited for mechanism transfer only.	analogous-domain	cited only, not redistributed	none	GRADE low health-system EHR default redesign study (before/after); non-randomised, different clinical decision (generics), cited for mechanism transfer only.	as cited in lib/tcoc/content/nudge-evidence.ts	2026-08-12	ARKA-TCOC /evidence/nudges /risk
Persell SD, Doctor JN, Friedberg MW, et al. Behavioral Interventions to Increase Statin Prescribing: A Cluster Randomized Clinical Trial. JAMA Cardiol. 2016. JAMA Cardiology Cluster-RCT of behavioural interventions on statin prescribing; cited for transfer of accountable-justification / peer-comparison mechanisms. GRADE — moderate: cluster-randomised, low risk of bias; downgraded one level for indirectness — statin prescribing, not imaging orders.	analogous-domain	cited only, not redistributed	none	GRADE moderate cluster-randomised, low risk of bias; downgraded one level for indirectness — statin prescribing, not imaging orders.	as cited in lib/tcoc/content/nudge-evidence.ts	2026-08-12	ARKA-TCOC /evidence/nudges /risk
Robinson NB, Gao M, Patel PA, Davidson KW, Peacock J, Herron CR, Baker AC, Hentel KA, Oh PS. Secondary review reduced inpatient MRI orders and avoidable hospital days. Clin Imaging. 2022 Feb;82:156–160. Clinical Imaging Independent academic authors (NewYork-Presbyterian / Weill Cornell Medicine and collaborators). No vendor funding disclosed in the abstract. PubMed record has no GrantList. COI class assigned from the abstract — full text not read. GRADE — low: observational pre-post quality-improvement design, no concurrent control group, no reported risk adjustment, two academic centres in one metro, effect not separable from secular trend. Effect size is not a reason to upgrade — the size of an effect and the certainty of an effect are different quantities.	imaging	cited only, not redistributed	none	GRADE low observational pre-post quality-improvement design, no concurrent control group, no reported risk adjustment, two academic centres in one metro, effect not separable from secular trend. Effect size is not a reason to upgrade — the size of an effect and the certainty of an effect are different quantities.	2022 Feb;82:156–160; PMID 34844100	2022-02-01	/evidence/nudges /risk ARKA-IP
Samuelson W, Zeckhauser R. Status quo bias in decision making. J Risk Uncertain. 1988;1:7–59. Journal of Risk and Uncertainty Foundational status-quo-bias paper; cited for behavioural mechanism framing on public nudge pages. GRADE — low: foundational behavioural-economics paper on status-quo bias; not an imaging intervention trial — certainty is for the mechanism claim only.	analogous-domain	cited only, not redistributed	none	GRADE low foundational behavioural-economics paper on status-quo bias; not an imaging intervention trial — certainty is for the mechanism claim only.	as cited in lib/tcoc/content/nudge-evidence.ts	2026-08-12	ARKA-TCOC /evidence/nudges /risk
Thaler RH. Mental accounting and consumer choice. Mark Sci. 1985;4(3):199–214. ("peanut effect" / mental accounting in incentive design). Marketing Science Thaler mental-accounting paper ("peanut effect"); cited for incentive-design framing on public nudge pages. GRADE — low: foundational mental-accounting paper; cited for incentive-design framing, not a graded clinical effect estimate.	analogous-domain	cited only, not redistributed	none	GRADE low foundational mental-accounting paper; cited for incentive-design framing, not a graded clinical effect estimate.	as cited in lib/tcoc/content/nudge-evidence.ts	2026-08-12	ARKA-TCOC /evidence/nudges /risk
Updating a Claims-Based Measure of Low-Value Services Applicable to Medicare Fee-for-Service Beneficiaries. J Gen Intern Med. 2022. CMS / HHS Independent peer-reviewed academic publication (or Cochrane review); not society-, vendor-, or payer-authored. Legacy licence text: `public-academic` ICD-10 carry-forward for the eleven LVI measure definitions. GRADE — low: observational methods paper on ICD-9→ICD-10 carry-forward for claims measures; not a randomised intervention effect — certainty is for coding continuity, downgraded for observational design and imprecision of GEM mapping.	imaging	cited only, not redistributed	none	GRADE low observational methods paper on ICD-9→ICD-10 carry-forward for claims measures; not a randomised intervention effect — certainty is for coding continuity, downgraded for observational design and imprecision of GEM mapping.	2022 JGIM	2026-08-07	ARKA-TCOC lvi-asympt-cardiac lvi-carotid-asympt lvi-headache-imaging lvi-lbp-imaging lvi-msk-early-mri lvi-preop-cxr lvi-preop-echo lvi-preop-stress lvi-repeat-ctap lvi-sinusitis-imaging lvi-syncope-imaging
Wagner / Fretheim / Penfold / Cochrane EPOC ITS guidance Cochrane Collaboration Independent peer-reviewed academic publication (or Cochrane review); not society-, vendor-, or payer-authored. Legacy licence text: Published academic / Cochrane EPOC GRADE — low: methodological standards for interrupted time series (Cochrane EPOC), not an intervention effect estimate; rated low as established methods guidance rather than a graded clinical effect.	methodological	cited only, not redistributed	none	GRADE low methodological standards for interrupted time series (Cochrane EPOC), not an intervention effect estimate; rated low as established methods guidance rather than a graded clinical effect.	As cited in Part A.8 (Wagner 12/12 default)	2026-08-07	ARKA-TCOC

government dataset (12)

Source	Domain	Licence	COI	GRADE	Version	Date	Used by
AHRQ CCSR (Clinical Classifications Software Refined) AHRQ Legacy licence text: `federal-public` Not applicable — this is a government claims dataset, not an effect estimate.	—	public domain	government	—	`v2026.1` (DXCCSR)	2026-08-07	ARKA-TCOC
AHRQ Elixhauser Comorbidity Software Refined AHRQ Legacy licence text: `federal-public` Not applicable — this is a government claims dataset, not an effect estimate.	—	public domain	government	—	Refined ICD-10-CM `v2026.1`	2026-08-07	ARKA-TCOC
CMS DE-SynPUF (Data Entrepreneurs’ Synthetic PUF) CMS / HHS Legacy licence text: `federal-public` (synthetic) Not applicable — this is a government claims dataset, not an effect estimate.	—	public domain	government	—	2008–2010 5% / 20 datasets	2026-08-07	ARKA-TCOC
CMS ordering/referring NPI enforcement (advanced imaging; PECOS; Item 17b / K3) CMS / HHS Legacy licence text: `federal-public` Not applicable — this is a government claims dataset, not an effect estimate.	—	public domain	government	—	Policy effective 2014-01-06; dictionaries as retrieved	2026-08-07	ARKA-TCOC
CMS-HCC V28 (payment year 2024+; full phase-in 2026) CMS / HHS Legacy licence text: `federal-public` Not applicable — this is a government claims dataset, not an effect estimate.	—	public domain	government	—	V28	2026-08-07	ARKA-TCOC
KFF — ACA Marketplace claims denials & appeals (2023) KFF Nonprofit policy research organisation analysis of Marketplace denial data; no financial interest in ARKA outputs. Not applicable — this is a government claims dataset, not an effect estimate.	—	public domain	none	—	as cited in lib/roi/roi-model.ts	2026-08-11	ROI
Medicare Physician & Other Practitioners — by Geography & Service CMS / HHS Legacy licence text: `federal-public` Not applicable — this is a government claims dataset, not an effect estimate.	—	public domain	government	—	CMS release year under ingest	2026-08-07	ARKA-TCOC
Medicare Physician & Other Practitioners — by Provider and Service (`data.cms.gov`) CMS / HHS Legacy licence text: `federal-public` Not applicable — this is a government claims dataset, not an effect estimate.	—	public domain	government	—	CMS release year under ingest	2026-08-07	ARKA-TCOC
MedPAC low-value care analyses (Schwartz measures applied to Medicare) CMS / HHS Legacy licence text: `federal-public` Primary anchor for measure definitions: lvi-asympt-cardiac, lvi-carotid-asympt, lvi-headache-imaging, lvi-lbp-imaging, lvi-msk-early-mri, lvi-preop-cxr, lvi-preop-echo, lvi-preop-stress, lvi-repeat-ctap, lvi-sinusitis-imaging, lvi-syncope-imaging. Not applicable — this is a government claims dataset, not an effect estimate.	—	public domain	government	—	Report year matched to measure-year under test	2026-08-07	ARKA-TCOC lvi-asympt-cardiac lvi-carotid-asympt lvi-headache-imaging lvi-lbp-imaging lvi-msk-early-mri lvi-preop-cxr lvi-preop-echo lvi-preop-stress lvi-repeat-ctap lvi-sinusitis-imaging lvi-syncope-imaging
MSSP / ACO Shared Savings Program PUFs U.S. federal government Legacy licence text: `federal-public` Not applicable — this is a government claims dataset, not an effect estimate.	—	public domain	government	—	Performance year under calibration	2026-08-07	ARKA-TCOC
Synthea (MITRE) + CMS RIF exporter CMS / HHS Open-source tooling used in ARKA's pipeline; authors have no financial interest in ARKA's clinical or utilization outputs. Legacy licence text: Open-source (Synthea licence as published) Not applicable — this is a government claims dataset, not an effect estimate.	—	open licence	none	—	Seed `SYNTHEA_GOLDEN_SEED` (`lib/tcoc/ingest/adapters/synthea.ts`); jar pinned at gen time	2026-08-07	ARKA-TCOC
NCSL Health Costs, Coverage and Delivery State Legislation database National Conference of State Legislatures Watchlist trigger state-ed-utilisation-legislation. Search 2026-08-23 (enacted only) for emergency department / ED / advanced imaging / CT / MRI / utilization review / utilisation review / prior authorization / utilization management. No enacted state statute requires utilisation review of advanced imaging in the emergency department. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	cited only, not redistributed	none	—	Enacted-only Power BI database, retrieved 2026-08-23	2026-08-23	policy-watchlist /compliance/regulatory-standing

regulatory document (19)

Source	Domain	Licence	COI	GRADE	Version	Date	Used by
HHS OIG — Medicare Advantage prior-authorization denials HHS Office of Inspector General Independent published source cited for ROI model figures; not redistributed. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	public domain	government	—	as cited in lib/roi/roi-model.ts	2026-08-11	ROI
HL7 FHIR R4 ServiceRequest HL7 International HL7 is a standards development organisation; FHIR ServiceRequest is an open standard with no financial interest in imaging utilization. Legacy licence text: Open HL7 FHIR Not applicable — this is a regulatory or policy document, not an effect estimate.	—	open licence	none	—	R4	2026-08-09	ARKA-AJ
LEAD (Long-term Enhanced ACO Design) Model CMS Innovation Center States LEAD is a 10-year voluntary model, 1 January 2027 through 31 December 2036; launches following the conclusion of ACO REACH at the end of 2026. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	public domain	government	—	Page last modified 2026-07-20	2026-07-20	ARKA-TCOC /lead /risk
Centers for Medicare & Medicaid Services Long-term Enhanced ACO Design (LEAD) Model FAQ CMS Innovation Center For ACOs entering in PY 2027, Base Years are CY 2024, CY 2025, and CY 2026; historical baseline stays fixed over the ACO’s participation (no rebasing within the ten-year horizon). Not applicable — this is a regulatory or policy document, not an effect estimate.	—	public domain	government	—	July 2026	2026-07-01	ARKA-TCOC /lead /risk /pricing expand-ii
LEAD Model Overview CMS Innovation Center LEAD model design. Ten-year voluntary ACO model beginning 1 January 2027; ten-year performance period without rebasing. The no-rebasing term is the load-bearing fact in the Expand II compounding / remaining-term willingness-to-pay model. Retrieved 19 August 2026. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	public domain	government	—	as published on cms.gov	2026-03-01	ARKA-TCOC /lead /pricing expand-ii
LEAD Model Payment Factsheet CMS Innovation Center Source for the two risk options (Global 100% / Professional 50%), the 3% quality withhold, the 1.75-3% Global explicit discount, the 10/30/60 base-year weighting, and the regional efficiency / prior savings adjustments. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	public domain	government	—	as published	2026-08-12	lead-benchmark lead-brief /lead risk roi
ACO REACH Model CMS Innovation Center ACO REACH concludes at the end of 2026; LEAD is the Innovation Center successor. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	public domain	government	—	as published on cms.gov	2026-07-20	ARKA-TCOC /lead /risk
WISeR (Wasteful and Inappropriate Service Reduction) Model CMS Innovation Center Model live 1 January 2026 in NJ, OH, OK, TX, AZ, WA; PA requests from 5 January 2026 for services on/after 15 January 2026. CMS: non-payment recommendations determined by appropriately licensed clinicians. Verified 2026-08-11 vs Provider & Supplier Operational Guide 4.0 — 13 active service categories; advanced imaging (diagnostic CT/MR/PET/NM) is not among them; Deep Brain Stimulation (NCD 160.24) and Percutaneous Image-Guided Lumbar Decompression (NCD 150.13) delayed. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	public domain	government	—	Page last modified 2026-08-04	2026-01-01	ARKA-INS never-auto-deny /governance/never-do /compliance/regulatory-standing
California Senate Bill 1120 — Physicians Make Decisions Act (Health & Safety Code §1367.01) California Legislature Effective 1 January 2025. Requires that a denial, delay, or modification of health care services based on medical necessity be made by a licensed physician or other health care provider competent to evaluate the clinical issues — not by an algorithm alone. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	public domain	government	—	Chapter 879, Statutes of 2024	2025-01-01	ARKA-INS never-auto-deny /governance/never-do
Calendar Year 2027 Hospital Outpatient Prospective Payment System (OPPS) and Ambulatory Surgical Center (ASC) Proposed Rule (CMS-1850-P) CMS Watchlist trigger cy2027-site-neutral-imaging. Proposed rule; site-neutral payment for imaging without contrast in excepted off-campus PBDs is not finalized. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	public domain	government	—	CMS-1850-P, issued 2 July 2026	2026-07-02	policy-watchlist /compliance/regulatory-standing
CMS Long-term Enhanced ACO Design (LEAD) Model FAQ — application window CMS Innovation Center Watchlist trigger lead-application-milestones. No application dates for a LEAD cohort after PY2027 are published in this FAQ. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	public domain	government	—	July 2026 FAQ	2026-07-01	policy-watchlist /compliance/regulatory-standing
ROOT Act (S.1692) — Radiology Outpatient Ordering Transmission Act U.S. Congress Watchlist trigger root-act-enacted stageHistory: introduced 2025-05-08 (Blackburn / Cortez Masto). Not enacted. Would revive a Medicare advanced-imaging AUC consultation mandate without the original claims-reporting mechanism. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	public domain	government	—	119th Congress, S.1692	2025-05-08	policy-watchlist /compliance/regulatory-standing
H.R.5737 — House companion / Patients First Act vehicle (Harshbarger / Moore) U.S. Congress Watchlist trigger root-act-enacted stageHistory: introduced 2025-10-10 (Harshbarger / Moore). House vehicle later used for Patients First Act packaging of ROOT provisions. Not enactment. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	public domain	government	—	119th Congress, H.R.5737	2025-10-10	policy-watchlist /compliance/regulatory-standing
House Patients First Act (H.R.5737) — ROOT provisions incorporated 15 July 2026 U.S. Congress Watchlist trigger root-act-enacted stageHistory: included_in_package 2026-07-15. A bill in a package is not a rule and is not demand. Enactment has not occurred. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	public domain	government	—	119th Congress, H.R.5737 — ROOT inclusion 2026-07-15	2026-07-15	policy-watchlist /compliance/regulatory-standing
CMS Shared Savings Program / LEAD attribution and benchmark methodology CMS Attribution and historical benchmarks for Shared Savings Program and LEAD ACOs are set by CMS programme rule (assignment methodology and LEAD payment/benchmark specifications), not by ARKA. ARKA consumes customer-supplied or CMS-rule assignments as read-only inputs. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	public domain	government	—	MSSP Shared Savings and Losses, Assignment, and Quality Performance Standard Methodology Specifications Version 14	2026-08-07	ARKA-TCOC /compliance/regulatory-standing no-attribution-or-risk-adjustment
Texas Insurance Code Chapter 4201, Subchapter N — preauthorization exemptions (HB 3459 as amended by HB 3812) Texas Legislature / Texas Department of Insurance Texas HB 3459 (87th Legislature, 2021) as amended by HB 3812 (89th Legislature, 2025). 90% approval-rate threshold (statute) over a 12-month evaluation period; fully insured TDI-regulated plans only — self-funded ERISA plans are out of scope. Companion TDI FAQ retrieved the same day: https://www.tdi.texas.gov/health/hb3459-faq.html. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	public domain	government	—	HB 3812 enrolled, effective 2025-09-01	2025-09-01	gold-card lib/goldcard/statutes.ts /ins/provider/gold-card
CMS Medicare Coverage Database — National Coverage Determinations Centers for Medicare & Medicaid Services Living NCD index. Version + retrieval date is the pin; a later MCD edit is not this source. The ncd adapter returns indeterminate unless sourced clinical criteria are on the order — it does not guess medical necessity. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	public domain	government	—	Medicare Coverage Database retrieval 2026-08-18 (IOM Pub 100-03)	2026-08-18	lib/coverage/adapters/ncd.ts lib/coverage/data/ncd-index.ts lib/coverage/registry.ts
CMS Medicare Coverage Database — Local Coverage Determinations Centers for Medicare & Medicaid Services LCDs differ by MAC. The mac-lcd adapter refuses to evaluate an order whose jurisdiction it does not know rather than falling back to a default LCD. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	public domain	government	—	Medicare Coverage Database retrieval 2026-08-18	2026-08-18	lib/coverage/adapters/mac-lcd.ts lib/coverage/data/mac-jurisdictions.ts lib/coverage/registry.ts
ONC Health Data, Technology, and Interoperability (HTI-1) — Decision Support Interventions source attributes, 45 CFR 170.315(b)(11) Office of the National Coordinator for Health Information Technology Voluntary ARKA publication of predictive DSI source attributes. ARKA is not a certified Health IT Module. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	public domain	government	—	HTI-1 final rule predictive DSI source attributes	2024-01-09	framework-crosswalk lib/standard/crosswalk.ts lib/governance/dsi-attributes.ts /trust /docs/model-card

professional society statement (8)

Source	Domain	Licence	COI	GRADE	Version	Date	Used by
AMA 2024 Prior Authorization Physician Survey American Medical Association AMA physician survey on prior authorization burden; AMA also licenses CPT. Not applicable — this is a professional society statement or survey, not an effect estimate.	—	cited only, not redistributed	society with a financial interest in the service	—	as cited in lib/roi/roi-model.ts	2026-08-11	ROI
AMA CPT® code set (identifiers only) American Medical Association Professional society that licenses the named code set or measure product; ARKA cites identifiers only and does not redistribute the corpus. Legacy licence text: Licensed identifiers cited only — do not embed CPT long descriptors or the codebook as a corpus; representative study codes used as FHIR Coding.code values Not applicable — this is a professional society statement or survey, not an effect estimate.	—	cited only, not redistributed	society with a financial interest in the service	—	CPT as cited in `MODALITY_TO_CPT`	2026-08-09	ARKA-AJ
CAQH Index (2023 / 2024) CAQH Multi-stakeholder industry index for administrative transaction costs; cited for PA cost/time figures only. Not applicable — this is a professional society statement or survey, not an effect estimate.	—	cited only, not redistributed	none	—	as cited in lib/roi/roi-model.ts	2026-08-11	ROI
Choosing Wisely (ABIM Foundation) ABIM Foundation ABIM Foundation Choosing Wisely campaign — recommendations cited; mark not used in ARKA UI. No financial interest in imaging utilization outcomes. Legacy licence text: Recommendations public; mark held by ABIM Foundation — cite, do not brand Not applicable — this is a professional society statement or survey, not an effect estimate.	—	cited only, not redistributed	none	—	Recommendation texts as cited per measure	2026-08-07	ARKA-TCOC lvs-homocysteine-screening lvs-t3-hypothyroid lvs-vitamin-d-screening
MGMA / Change Healthcare — cost to rework a denied claim MGMA Professional society practice survey / cost figures; cited for administrative rework cost only. Not applicable — this is a professional society statement or survey, not an effect estimate.	—	cited only, not redistributed	society with a financial interest in the service	—	as cited in lib/roi/roi-model.ts	2026-08-11	ROI
Patel MS, Volpp KG, Asch DA. Nudge Units to Improve the Delivery of Health Care. N Engl J Med. 2018;378(3):214–216. New England Journal of Medicine NEJM perspective on nudge units in health care delivery; cited for institutional framing, not as a graded intervention effect. Not applicable — this is a professional society statement or survey, not an effect estimate.	—	cited only, not redistributed	none	—	as cited in lib/tcoc/content/nudge-evidence.ts	2026-08-12	ARKA-TCOC /evidence/nudges /risk
Thaler RH, Sunstein CR. Nudge: Improving Decisions About Health, Wealth, and Happiness. Yale University Press; 2008. Yale University Press Popular book synthesizing behavioural-economics literature (defaults / status quo); cited for framing on public nudge pages, not as an ARKA effect estimate. Not applicable — this is a professional society statement or survey, not an effect estimate.	—	cited only, not redistributed	none	—	as cited in lib/tcoc/content/nudge-evidence.ts	2026-08-12	ARKA-TCOC /evidence/nudges /risk
Joint Commission and CHAI — Responsible Use of AI in Healthcare (RUAIH) guidance The Joint Commission / Coalition for Health AI RUAIH element titles used as crosswalk field names. ARKA is not Joint Commission certified for Responsible Use of AI. Not applicable — this is a regulatory or policy document, not an effect estimate.	—	cited only, not redistributed	none	—	Initial guidance 17 September 2025; voluntary certification announced May 2026	2025-09-17	framework-crosswalk lib/standard/crosswalk.ts lib/governance/ai-package.ts /trust /governance/ai

technical specification (5)

Source	Domain	Licence	COI	GRADE	Version	Date	Used by
HL7 Da Vinci Coverage Requirements Discovery (CRD) Implementation Guide HL7 International / Financial Management Work Group (Da Vinci Project) Pinned by docs/DAVINCI_CONFORMANCE_DECISION_2026-09.md. Package hl7.fhir.us.davinci-crd 2.2.1 (package.manifest.json date 20260327113228). Permanent STU home: https://hl7.org/fhir/us/davinci-crd/2.2.1/. Not applicable — this is a technical specification, not an effect estimate.	—	cited only, not redistributed	none	—	2.2.1	2026-03-27	docs/DAVINCI_CONFORMANCE_DECISION_2026-09.md lib/coverage/types.ts lib/coverage/adapters/davinci-dtr.ts lib/davinci/crd.ts
HL7 Da Vinci Documentation Templates and Rules (DTR) Implementation Guide HL7 International / Financial Management Work Group (Da Vinci Project) Pinned by docs/DAVINCI_CONFORMANCE_DECISION_2026-09.md. Package hl7.fhir.us.davinci-dtr 2.2.0 (package.manifest.json date 20260327160408). Permanent STU home: https://hl7.org/fhir/us/davinci-dtr/2.2.0/. Not applicable — this is a technical specification, not an effect estimate.	—	cited only, not redistributed	none	—	2.2.0	2026-03-27	docs/DAVINCI_CONFORMANCE_DECISION_2026-09.md lib/coverage/types.ts lib/coverage/adapters/davinci-dtr.ts lib/davinci/dtr.ts
HL7 Da Vinci Prior Authorization Support (PAS) Implementation Guide HL7 International / Financial Management Work Group (Da Vinci Project) Pinned by docs/DAVINCI_CONFORMANCE_DECISION_2026-09.md. Package hl7.fhir.us.davinci-pas 2.2.1 (package.manifest.json date 20260327135848). Permanent STU home: https://hl7.org/fhir/us/davinci-pas/2.2.1/. Not applicable — this is a technical specification, not an effect estimate.	—	cited only, not redistributed	none	—	2.2.1	2026-03-27	docs/DAVINCI_CONFORMANCE_DECISION_2026-09.md lib/coverage/types.ts lib/coverage/adapters/davinci-dtr.ts lib/davinci/pas.ts
ARKA-INS coverage adapter (clinical engine wrap) ARKA Wraps determineCoverage() in lib/ins/coverage-adapter.ts as one CriteriaSource among five. Records a caller-supplied determination; it does not look up a clinical rulebook. Not applicable — this is a technical specification, not an effect estimate.	—	not licensed — referenced by citation	vendor-authored	—	ins-coverage-adapter-1	2026-08-18	lib/ins/coverage-adapter.ts lib/coverage/adapters/arka-clinical.ts lib/coverage/registry.ts
CHAI Applied Model Card schema v0.2 Coalition for Health AI (CHAI) Open-source Applied Model Card schema. Vendored XSD at docs/chai/applied-model-card-v0.2.xsd. Field names used verbatim; not a claim of CHAI endorsement. Not applicable — this is a technical specification or instrument, not an effect estimate.	—	open licence	none	—	v0.2	2024-11-22	framework-crosswalk lib/standard/crosswalk.ts lib/standard/chai-model-card.ts /trust /api/standard/model-card

evidence panel record (1)

Source	Domain	Licence	COI	GRADE	Version	Date	Used by
ARKA Evidence Panel record epr-2026-08-23-cardiac-preop-arrival ARKA Standing Evidence Panel Sealed panel record (6 decision(s); charter 1.0.0). Cite JSON or PDF at the stable API URL. Not applicable — an evidence panel record attests to a certainty adjudication; it is not primary clinical evidence for the underlying claim (Prompt 33.7).	—	public domain	none	—	sha256:bbcd860ddbdf00f202db9fcf8df4e4b915caaa8cc3bb1eed716efefb7970275f	2026-08-23	lib/governance/evidence-panel/record.ts app/governance/evidence-panel/page.tsx lib/aiie/knowledge-matrix

Pending classification

No sources are currently held out of the register. When a new external source cannot be mapped to licence and COI without guessing, it appears here instead of being invented.

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.