Clinical documentation and workforce support AI

Reviews and evidence

• Official Suki Assistant scope source Vendor evidence · Verified source

  The supplier page is used to anchor what Suki Assistant publicly says it does. It is a scope source, not independent proof of performance, safety, value, or local readiness.

  Open the source ->
• JMIR McLeod Health multiphase pilot study Independent review · Verified source

  The peer-reviewed McLeod Health study describes a structured four-vendor evaluation, live clinical simulations, EHR integration testing, a 90-day pilot, and system-wide implementation of the selected ambient AI solution. The paper acknowledges Suki staff contributions and says the authors independently verified summary extracts against internal data.

  Why this matters: The most useful evidence here is the selection method, not the headline result: an enterprise buyer can copy the staged evaluation, live scripts, integration test, and baseline measurement before choosing a supplier.

  Reviewer context

  Bryon Kenneth Frost, MD, Victor Eugene Collier, MD, Franklin Sturgill, Jessie Polson, and Jennifer Jones; the paper acknowledges Suki staff for summary extracts and describes independent verification against internal EHR and revenue data. Peer-reviewed health-system clinical and information-technology authors at McLeod Health.

  Organisation context

  McLeod Health is described as a nonacademic system with seven hospitals and more than 1,200 providers serving 18 counties in North and South Carolina. Size basis: The article states seven hospitals and more than 1,200 providers, supporting a regional health-system context rather than a global-enterprise size claim.

  Scope and sentiment

  exact product scope; positive signal; vendor involvement disclosed.

  Source trust

  4/5. Peer review, named health-system authors, disclosed vendor assistance, objective workflow testing, and internal-data verification are strong signals; the single-system design and vendor-supported extracts limit external transfer. 0.68 context weight.

  Implementation context

  Four vendors were tested with 15 complex outpatient scripts; the finalists were tested for Epic workflow integration; a 90-day pilot ran across five specialties before system-wide rollout. The paper reports early adoption and outcome measures but not a randomized comparison.

  Open the source ->
• KLAS cross-organisational ROI validation Independent review · Verified source

  KLAS names three health systems that deployed Suki and describes independent quantitative and qualitative ROI validation across clinician time, documentation, coding, operations, and finances. The report is linked from the vendor site, so the methodology and commercial relationship should be reviewed directly before relying on any metric.

  Why this matters: It adds cross-organisation evidence and measurement questions beyond a single customer story, while making the buyer inspect how ROI was attributed and whether the same EHR and workflow conditions apply.

  Reviewer context

  Mac Boyter, Sidnee Wood, and Tyson Blauer are named as KLAS report authors on the report page. Healthcare technology research and ROI analysts.

  Organisation context

  The report covers FMOL Health, McLeod Health, and Rush University System for Health, described as three large Epic-based health systems using Suki in ambulatory and emergency settings. Size basis: The report identifies three large health systems; it does not provide a single comparable workforce band for all three, so the band describes the operating context conservatively.

  Scope and sentiment

  exact product scope; positive signal; not disclosed.

  Source trust

  4/5. Named independent analysts and a multi-organisation methodology strengthen the signal, but the report page is vendor-hosted and the full report method, sample boundaries, and commercial relationship need direct review. 0.80 context weight.

  Implementation context

  The report states that KLAS evaluated quantitative and qualitative ROI, longitudinal Epic Signal data, clinician burden, documentation quality, coding accuracy, operational efficiency, financial performance, governance, pricing, and rollout lessons.

  Open the source ->
• Citizens Memorial Health implementation case Customer story · Verified source

  The named rural health system reports MEDITECH Expanse integration, adoption across five specialties, 1,500 ambient notes in two months, and reduced after-hours documentation. These figures are vendor-published and are retained as a reference-call lead rather than a transferable forecast.

  Why this matters: It is a relevant nonacademic implementation reference for a buyer with MEDITECH or a rural operating model, and it exposes the adoption and measurement questions a reference call should test.

  Reviewer context

  Dr. Rusty Davis, nephrologist at Citizens Memorial Hospital, is quoted; the case identifies Citizens Memorial Health as the customer organisation. Named practising physician quoted in a vendor-published customer case.

  Organisation context

  Citizens Memorial Health is presented as a rural health system; the public case describes MEDITECH Expanse and five specialties but does not state a workforce size. Size basis: The page identifies the rural health organisation and deployment footprint but does not publish a comparable employee or clinician count.

  Scope and sentiment

  exact product scope; positive signal; vendor published.

  Source trust

  3/5. A named clinician, named customer, product integration, and concrete measures are useful primary evidence, but the case is vendor-published and the method is not independently audited. 0.36 context weight.

  Implementation context

  The page reports 80% adoption across five specialties, 1,500 notes in two months, and a 41% reduction in after-hours documentation; the baseline and measurement method are not independently audited on the page.

  Open the source ->

Public product visual references

Reviews and evidence

• Official Abridge scope source Vendor evidence · Verified source

  The supplier page is used to anchor what Abridge publicly says it does. It is a scope source, not independent proof of performance, safety, value, or local readiness.

  Open the source ->
• JAMA Network Open multi-health-system study Independent review · Verified source

  A peer-reviewed quality-improvement study evaluated the same ambient AI scribe across six US health systems. Abridge AI involvement is disclosed in the author affiliations and methods; the reported outcomes are not treated as a universal product result.

  Why this matters: It is stronger than a vendor testimonial because the population, workflow, measures, and limitations are visible, while the design still requires a buyer to reproduce the result in its own clinical workflow.

  Reviewer context

  Kristine D. Olson, MD, MSc, Daniella Meeker, PhD, Matt Troup, PA-C, and coauthors; the article discloses an Abridge AI affiliation and describes independent analysis roles for Olson and Meeker. Peer-reviewed health-policy researchers and clinical leaders across six health systems.

  Organisation context

  The study included 263 physicians and advanced practice practitioners from six academic and community health systems, with 451 clinicians enrolled and 30 days of use. Size basis: The study is multi-site and reports participant counts rather than a comparable workforce or revenue band for each health system.

  Scope and sentiment

  exact product scope; positive signal; vendor involvement disclosed.

  Source trust

  4/5. Peer review, named authors, disclosed vendor involvement, multi-site data, and explicit methods make this strong evidence; the quality-improvement design, voluntary participation, self-reported outcomes, and vendor involvement limit causal transfer. 0.48 context weight.

  Implementation context

  Clinicians obtained patient consent, recorded ambulatory encounters, reviewed and edited generated notes, could inspect transcript or audio, and imported the final note into the EHR; the study reports pre/post survey outcomes rather than a randomized control.

  Open the source ->
• UVM Health Network enterprise case study Customer story · Verified source

  The named academic health system reports an evaluation of multiple ambient documentation solutions, a 50-provider primary-care evaluation, and enterprise rollout. Reported impact is vendor-published and must be checked against the case methodology and buyer baseline.

  Why this matters: It gives an enterprise buyer a named reference and a concrete set of questions about evaluation, onboarding, note quality, and wellbeing without turning a customer case into a promised outcome.

  Reviewer context

  Jason Sanders, CEO and President of the UVM Health Network Medical Group, and named clinicians Alicia Jacobs, MD, Marie Sandoval, MD, and Sean Maloney. Health-system executive and practising clinician voices quoted in a vendor-published customer case.

  Organisation context

  The University of Vermont Health Network is described as an integrated academic health system serving more than one million people, with 1,100 clinicians. Size basis: The public case states 1,100 clinicians and a health-system footprint across Vermont and northern New York.

  Scope and sentiment

  exact product scope; positive signal; vendor published.

  Source trust

  3/5. Named executives, clinicians, organisation scale, and a described evaluation are useful primary evidence, but the source is vendor-published and the underlying baseline, comparator, and statistical method are not independently audited on the page. 0.60 context weight.

  Implementation context

  The case says 50 primary-care providers were involved in evaluation and that the Digital and Remote Health Committee selected Abridge for enterprise rollout; the reported 53% professional-fulfillment and 60% after-hours-documentation figures are vendor-published.

  Open the source ->

Public product visual references