A²I™ Device Corpus
Drawn from published MSAC, NICE, and PHARMAC device guidance, plus real CMS National Coverage Determinations (United States Medicare — the largest single US public funding channel, not the only one; Medicaid and private payers aren't covered here). This is pattern-level intelligence, not yet a full per-decision corpus for every body — the next phase sources further individual historical device funding decisions to match the depth of the 158-decision PTAC pharma corpus.
Clinical domains carry the primary evidence burden, funding domains cover the economic case, and Equity (DD10) is a non-overridable gate: if there is no real evidence for how a device reaches underserved or high-need populations, the overall score is hard-capped regardless of every other domain. The weight shown is each domain's deterministic Impact input to Claim Weight (ACS) — never a freehand number.
Does the device change clinical management in a way that improves patient outcomes — not merely that it functions correctly or is safe.
Adverse event evidence, long-term safety data for implantables, post-market surveillance and registry requirements.
Evidence against the correct standard-of-care comparator for the target market — not placebo, not an international comparator, not the current funded alternative left unaddressed.
Single-arm vs. controlled design, pre-specification of the primary endpoint, surrogate vs. patient-relevant endpoints, sample size adequacy.
Whether the evidence population bridges to the target market's clinical practice, disease prevalence, and treatment pathway.
Sufficiency of follow-up duration to capture clinically meaningful, durable outcomes, and real-world evidence following initial deployment.
Presence and adequacy of a health economic model meeting the target body's reference case (QALY/cost-utility, Australian/UK/NZ-specific costs and epidemiology).
Operational and budget-impact evidence: site-of-service requirements, operator-volume/team requirements, training burden, supply chain and implementation suitability.
Coverage-pathway strategy: Coverage with Evidence Development plans, Patient Access Scheme-equivalent proposals, national vs. fragmented local coverage strategy, shared decision-making documentation for preference-sensitive devices.
Equity-of-access evidence for underserved/high-deprivation populations (e.g. Aboriginal and Torres Strait Islander, rural/remote, Māori and Pacific). Non-overridable gate: if this domain has zero confirmed evidence-quality findings, the composite score is hard-capped at 2.5 regardless of every other domain.
Unlike the patterns above, these are named, individually-sourced device decisions — a real device, a real committee or agency outcome, and a direct citation to the published PSD, NICE guidance, or CMS coverage determination. CMS is included here (not FDA) because it is the actual US body that decides device funding — FDA only decides market approval, a different question covered separately as reference material. PHARMAC individual decisions are not yet included. This is a representative sample built by locating and reading individual decisions one at a time, not exhaustive coverage of any body's full history.
published 8 March 2021
Not recommended for routine NHS adoption. Further research required before reconsideration.
published 14 October 2020
Not recommended for routine adoption. A randomised controlled trial against standard risk assessment was specified as the evidence needed to reconsider.
published 11 April 2023
Recommended only for a defined subgroup — patients who meet knee-replacement referral criteria but decline surgery — conditional on real-world data collection. Explicitly not recommended for patients unfit for surgery, pending further research.
published 3 September 2020
Recommended for adoption, positioned as an option after conservative and drug treatment fail.
published 3 December 2019
Recommended for adoption, but with a mandatory stopping rule: only patients who respond within the first 3 months continue treatment.
published 6 January 2022
Adopted, but NHS supply was paused between December 2022 and July 2023 until the manufacturer met the Digital Technology Assessment Criteria (DTAC) — illustrating that a positive recommendation is not necessarily a permanent, unconditional funding outcome.
published 29 November 2011
Supported for MBS funding, but only for single-level cervical degenerative disc disease in skeletally mature patients with a stable spine, no prior cervical surgery, who had failed conservative therapy.
published 29-30 November 2012
Not supported for MBS funding.
published 1-2 April 2015
Not supported for MBS funding, resubmission invited.
published 30-31 March 2016
Supported for MBS funding, conditional on a fee cut, mandatory heart-team assessment of every patient, facility/operator accreditation, and mandatory outcome-registry reporting.
published 29-30 July 2021
Supported — new MBS item created, device-agnostic descriptor, extending TAVI funding to low-risk patients.
published 29-30 July 2021
Deferred rather than confirmed or withdrawn — MSAC asked for the funded population to be redefined and the economic model rebuilt before it would give a clean recommendation.
published 21 June 2019 (v2; original 1 May 2012)
Covered for symptomatic aortic valve stenosis within an FDA-approved indication, contingent on Coverage with Evidence Development (CED) requirements.
published 8 February 2016
Covered for non-valvular atrial fibrillation only under CED, as a second-line alternative to anticoagulation.
published 1 January 2002
Covered for urge incontinence, urgency-frequency syndrome, and urinary retention, conditional on treatment history and a successful test-stimulation phase.
published 4 April 2005 (v2; superseded 26 September 2022)
Covered for bilateral moderate-to-profound sensorineural hearing loss in patients who gain limited benefit from hearing aids.
published 1 July 1999
Covered for medically refractory partial-onset seizures where surgery is not recommended or has failed.
published 4 May 2007
Not covered for treatment-resistant depression, despite FDA approval of the device for this indication in 2005.
published 1 December 2020
Recommended as an option for suspected arrhythmias needing >24h ambulatory ECG, conditional on NHS organisations collecting resource-use and longer-term outcome data.
published 11 February 2020
Shows promise but not enough evidence for routine NHS adoption; further research recommended.
published June 2014 (withdrawn April 2020; under review August 2024)
Originally recommended; guidance later withdrawn. After UK suspension of modified MAGEC X was lifted (Aug 2024), NICE prioritisation board is reviewing whether to reinstate or replace the guidance.
published 22 August 2012
May offer advantages for selected patients (e.g. fragile or damaged skin); comparative clinical evidence against disposable pads is limited.
published 31 January 2019
Recommended as cost-saving for diabetic foot ulcers and venous leg ulcers; not enough evidence for routine adoption in non-venous leg ulcers.
published 10 September 2018
Case for adopting Neuropad not supported by the evidence.
published 7 March 2022
Not recommended as a cost-saving option for diabetic foot ulcers.
published 8 March 2022
More research recommended for chronic wounds; not recommended for acute wounds because evidence is very limited.
published 31 March 2022
Potential patient and system benefits acknowledged, but more comparative research required before routine adoption.
published 30 November 2021
Shows promise for BCG-refractory / BCG-intolerant high-risk NMIBC, but not enough good-quality evidence for routine adoption; special arrangements only.
published 24 June 2020
Evidence supports adopting Rezum for moderate-to-severe LUTS with a moderately enlarged prostate; cost modelling estimated savings versus TURP/HoLEP.
published 23 January 2019
Case for adopting Senza HF10 SCS supported — at least as effective as low-frequency SCS without paraesthesia.
published 13 February 2017
Case for adopting HeartFlow FFRCT supported for stable recent-onset chest pain offered CCTA (64-slice or above).
published 4 May 2021
Evidence supports adopting UroLift as a minimally invasive alternative to TURP/HoLEP; modelling suggested cost savings.
published 18 January 2017
Covered for bradyarrhythmia under Coverage with Evidence Development (CED).
published Updated through Dec 2020 NCD Manual revisions
Covered for advanced heart failure meeting clinical and device-approval criteria (including destination-therapy pathways).
published 15 February 2018
Covered for specified ventricular-arrhythmia / cardiomyopathy indications; shared decision-making required for some groups.
published 31 July 2023 (current version)
Covered under CED for symptomatic moderate-to-severe/severe functional MR despite GDMT, or significant symptomatic degenerative MR per FDA-approved indication.
published 13 August 2013
Covered for non-reversible symptomatic bradycardia due to sinus node dysfunction or second/third-degree AV block; other indications left to MAC discretion.
published 1 April 2003
Covered for FDA-approved DBS targets in essential tremor and Parkinson’s disease when clinical and facility criteria are met.
published 7 August 1995
Covered for chronic intractable pain under late-resort criteria, including temporary electrode trial before permanent implantation.
published 17 December 2004 (CSII criteria)
Covered in the home setting for diabetic patients meeting C-peptide/autoantibody and intensive insulin-therapy criteria.
published Local coverage determination (MAC)
Covered under local MAC criteria for chronic intractable pain — not a national NCD.
published Local coverage determination (MAC)
Covered under local MAC criteria for diabetes management — not a national NCD.
published 28–29 November 2024
Supported — therapeutic technology; new MBS item pathway.
published 28–29 November 2024
Supported (therapeutic technology) for destination-therapy LVAD use.
published 28–29 July 2022
Supported for new MBS items covering leadless pacemaker procedures.
published 3–4 April 2025
Not supported.
published 23–24 November 2023
Supported.
published 30–31 March 2023
Not supported.
published 30–31 July 2015
Not supported for MBS funding.
published 1–2 April 2015
Not supported.
published 31 March–1 April 2022
Partially supported.
published 28 November 2013 (with later related meetings)
Supported.
published 28–29 November 2019
Not supported.
published 26–28 November 2014
Not supported.
published 27 July 2017
Supported.
published 25–26 November 2021
Not supported.
published 28–29 July 2022
Supported new MBS items for TUWA and PUL; retained EEP/VLAP; advised removing TUMT.
published 31 July–1 August 2025
Supported (therapeutic technology) for paediatric and adult acute burns.
DD4 · All, Diagnostic, Implantable, Digital Health / SaMD
MSAC consistently rejects clinical utility claims based on single-arm studies. Comparative evidence against current Australian standard of care is required. Single-arm studies are only accepted for safety and feasibility claims, not for clinical utility or cost-effectiveness.
DD3 · All
The comparator must reflect current Australian clinical practice — not placebo, historical control, or international standard of care unless explicitly justified. MSAC has deferred multiple applications where the comparator was not the Australian standard of care.
DD4 · Diagnostic, IVD, Digital Health / SaMD
MSAC requires patient-relevant outcomes — survival, quality of life, functional status, or validated intermediate endpoints with demonstrated linkage to patient outcomes. Diagnostic accuracy alone is insufficient without evidence of downstream patient benefit.
DD5 · All
Studies conducted entirely in non-Australian populations require explicit bridging justification. MSAC has rejected or deferred applications where prevalence, treatment pathway, or clinical setting differed materially from Australian practice.
DD7 · All
MSAC requires a full health economic model for all MBS listing applications, using Australian costs and epidemiology, presenting cost per QALY or cost per clinical outcome. Absence of an economic model is an automatic rejection driver.
DD6 · Implantable, All
MSAC has deferred applications where follow-up was insufficient to capture clinically meaningful outcomes. Implantable devices require a minimum 12 months; chronic condition devices require follow-up commensurate with the disease course.
DD10 · All
MSAC requires explicit consideration of access for Aboriginal and Torres Strait Islander peoples and rural/remote populations. Applications that do not address equity of access are flagged, weighted heavily where the condition has disproportionate burden in these populations.
DD1 · Diagnostic, IVD
MSAC separates assessment of safety, effectiveness, and cost-effectiveness. A device shown to be safe and to perform as claimed has not thereby shown clinical utility — it must show it changes clinical management in a way that improves patient outcomes.
DD4 · All
MSAC has rejected applications where sample sizes were insufficient to detect a clinically meaningful difference. Studies powered for regulatory clearance are frequently underpowered for MSAC's comparative-effectiveness purposes.
DD4 · All
Post-hoc endpoint selection or composite endpoints without pre-specification are a recurring MSAC concern. The primary endpoint must be pre-specified, clinically relevant, and powered accordingly.
DD1 · Digital Health / SaMD
NICE's Evidence Standards Framework (ESF) sets tier-specific requirements for digital health technologies. Tier A (self-care) needs acceptability/usability evidence; Tier B (guided self-care) needs effectiveness evidence; Tier C (treatment/clinical pathway) needs RCT-level or equivalent evidence. Claiming a higher tier than the evidence supports is a primary rejection driver.
DD7 · All
NICE requires a cost-utility analysis using QALYs, reference case threshold £20,000–£30,000/QALY (higher for end-of-life criteria: survival under 24 months, QALY gain ≥3 months). Applications without a QALY model are rejected at assessment stage.
DD5 · All
Evidence from non-UK populations must be explicitly bridged to UK NHS clinical practice, addressing differences in care pathways, disease prevalence, and treatment standards. US evidence in particular frequently requires substantial bridging.
DD4 · Diagnostic, IVD, Digital Health / SaMD
NICE requires validated surrogate endpoints — evidence that the surrogate reliably predicts the patient-relevant outcome. Unvalidated surrogates lead to high uncertainty ratings and typically a not-recommended or research-only outcome.
DD10 · All
Under the Equality Act 2010, NICE requires applicants to demonstrate consideration of impact on protected characteristics (age, sex, disability, ethnicity, etc). Missing equality impact assessments are a compliance gap that delays or blocks appraisal.
DD2 · Implantable
For implantable devices, NICE and the MHRA require long-term safety evidence — short-term studies are insufficient. Implant registries, post-market surveillance data, and adverse event reporting are expected.
DD6 · Digital Health / SaMD
For SaMD and digital health technologies, NICE increasingly requires real-world evidence of effectiveness in NHS settings following initial deployment. Bench performance and trial data alone are insufficient for sustained commissioning.
DD7 · All
NICE's reference case specifies perspective (NHS and personal social services), time horizon (lifetime), discount rate (3.5% costs and outcomes), and utility measurement (EQ-5D). Deviations without justification result in model rejection.
DD9 · All
Where cost-effectiveness is uncertain or the ICER exceeds threshold, NICE expects a Patient Access Scheme proposal. Absence of a PAS when the base case ICER is above £30,000 is a negative signal.
DD3 · Therapeutic / Surgical, Implantable
For devices in the treatment pathway (ESF Tier C), NICE requires comparative evidence. Single-arm studies demonstrating device performance are insufficient to establish comparative effectiveness against current NHS standard of care.
DD3 · All
PHARMAC's Factors for Consideration require comparison against the current funded alternative (an existing Schedule device or standard hospital practice), not an unfunded or international comparator, unless a case is made for a first-in-class listing.
DD7 · All
PHARMAC's investment approach weighs health need, health benefit, cost, and suitability within a fixed device/pharmaceutical budget. Applications without a budget impact analysis and cost-utility case are difficult to prioritise against competing investments.
DD10 · All
Under Te Tiriti o Waitangi obligations, PHARMAC requires explicit consideration of equity impact for Māori and Pacific peoples. Absent equity evidence can block funded access or generate binding conditions, consistent with the pattern seen across PHARMAC's pharmaceutical decisions.
DD5 · All
Evidence generated entirely offshore (commonly US, European, or Australian trials) requires bridging to the New Zealand health system, clinical pathway, and population profile before PHARMAC will treat it as directly applicable.
DD1 · All
PHARMAC's Factors for Consideration weigh the health need of the population against the strength and directness of clinical benefit evidence. Surrogate-only or indirect evidence weakens the case relative to devices with direct, patient-relevant outcome data.
DD8 · Implantable, Therapeutic / Surgical, Digital Health / SaMD
PHARMAC's suitability criterion considers practical implementation — training burden, supply chain, compatibility with existing hospital equipment/workflow. Devices with no NZ implementation or training plan face additional funding risk.
DD9 · Implantable, All
For novel or high-uncertainty devices, CMS frequently covers only under Coverage with Evidence Development — a national registry tracking safety/effectiveness outcomes for a defined period — rather than issuing an unconditional yes or no. Devices with no post-market data-collection plan face a higher risk of CED-only or non-coverage, as seen with TAVR and LAAC.
DD8 · Implantable
CMS has repeatedly conditioned coverage for higher-risk implantables on multidisciplinary team evaluation (e.g. TAVR's heart-team requirement), facility accreditation, and operator case-volume minimums. Applications silent on these operational conditions are less likely to secure unconditional coverage.
DD1 · All
CMS's coverage standard is independent of FDA clearance/approval — a device can be legally marketable and still be found not reasonable and necessary for Medicare if outcome evidence for the specific population is lacking. CMS declined coverage for vagus nerve stimulation in treatment-resistant depression on exactly this basis, despite FDA approval for that indication.
DD9 · All
Many device categories have no National Coverage Determination at all, leaving coverage to Local Coverage Determinations set independently by each Medicare Administrative Contractor. A device can be funded in one region and denied in another with no single national resolution — a fragmentation risk that a funding strategy should explicitly address.
DD3 · Implantable, Therapeutic / Surgical
Where a device proposes to replace or supplement a pharmaceutical standard of care (e.g. an anticoagulation alternative), CMS coverage has required evidence on appropriateness for patients who can and cannot tolerate the drug-based comparator long-term — not device performance data alone.
DD5 · All
CMS coverage often applies a specific clinical threshold (a risk score, symptom severity band, or audiometric cutoff) with a narrower, trial-only pathway for patients just outside it — as seen in cochlear implantation's borderline benefit-score carve-out. Evidence plans with no defined threshold or borderline-population strategy face ambiguous coverage at the margins.
DD9 · Implantable
For elective, preference-sensitive procedures with a viable non-device alternative, CMS has made a documented shared decision-making conversation between clinician and patient a coverage precondition, not just a best-practice recommendation.
DD9 · All
CMS's 2024 Transitional Coverage for Emerging Technologies (TCET) pathway offers expedited Medicare coverage for FDA-designated Breakthrough Devices, but requires manufacturer self-nomination roughly 12 months ahead of expected FDA action. Breakthrough-designated devices with no CMS engagement plan risk a coverage gap after FDA clearance.