PaceMate

The company frames its triage as the right kind of AI, and describes it as rules. The home page contrasts "a version of AI that takes clinical judgment out of the equation" with the company's "clinician-led intelligence", calls Auto-Triage a "proprietary algorithmic intelligence engine" that ranks every patient in a population by clinical urgency every day, and reports 61 percent less physician alert burden. The company's own August 2026 post on AI in cardiac remote monitoring states that "rules-based triage in our platform reduced alert burden by 61 percent in peer-reviewed work covering 132,000 patients", closing routine cases and routing exceptions to clinicians with longitudinal context: prior transmissions, arrhythmia trends, lead impedance and medication changes. A rule set is not a learned component, however much context it assembles.

Most of what PaceMate does needs no model: it aggregates discrete data from cardiac implantable devices (14 integrated manufacturers), pairs it with record data, generates reports and automates billing documentation through Auto-Script. Consumer ECG analysis brought in through AliveCor's KardiaPro platform (a 2023 integration) is AliveCor's own cleared software, not a PaceMate component. The company announced a head of AI to prepare the platform for AI expansion, which describes plans, not a product.

The marketing speaks the language of AI, while the triage it describes is rules based.

The autonomy is an alert filter and two parts of its oversight are missing. Auto-Triage prioritizes and filters device alerts, and the company reports a 61 percent reduction in physician alert burden; what does surface is reviewed and signed by a clinician through one click sign and next.

The company publishes neither the criteria or clinical guideline the triage applies nor its sensitivity or false negative rate, and it does not say what happens to filtered transmissions (whether they stay retrievable, auditable or sampled for review). With both the threshold and the after the fact check missing, the control on the alerts nobody sees is asserted rather than described. Auto-Script also initiates billing documentation and missed charge recovery, with no stated human review before a claim is submitted.

The workflow is documented in detail, and the triage approach is named. The company names its remote monitoring procedure codes, 31 and 91 day billing cycles, bidirectional synchronization with Epic (SMART on FHIR verified), Oracle Health and athenahealth, discrete device data capture, a five step workflow and one click sign and next. Its August 2026 post describes Auto-Triage as rules based triage that closes routine transmissions and routes exceptions to clinicians with longitudinal context attached (comparison with prior transmissions, arrhythmia trends, lead impedance and medication changes), and cites peer reviewed work covering 132,000 patients with 61 percent less overall alert burden and 30 percent less physician alert burden.

The rules themselves are not published: no criteria for closing a transmission, sensitivity for events the rules close, version or change practice is described. Consumer ECG analysis brought in through AliveCor's KardiaPro platform is AliveCor's.

The approach and its inputs are named, while how the rules are versioned and updated is not described.

Partners are named and the platform's own model layer is not. AliveCor's KardiaPro platform was integrated in 2023 to bring consumer ECG data and its analysis into PaceMate, Medtronic's SessionSync supports in clinic wireless workflow, and Epic, Oracle Health and athenahealth are named record system integrations; the privacy policy names Google Analytics and refers to vendors and service providers without naming them.

Nothing says whether Auto-Triage is built in house, licensed or derived from device manufacturer logic, no hosting provider or subprocessor list is published, and the research grade dataset the company says outside parties use sends data to recipients who are not named.

Named deployments with dated operational results, and research that mostly evaluates other companies' algorithms. The home page reports more than 2.3 million patients managed and 33 peer reviewed publications, a 61 percent reduction in physician alert burden from Auto-Triage (presented at HRS 2026), 543 hours saved annually at Ascension Jacksonville, Piedmont Healthcare reaching its two year billing goal in under a year, 75 percent average billing capture within 90 days of go live, 98 percent less manual billing entry and 70 percent less documentation time.

A Heart Rhythm study (2024) used de-identified data from 26 US centers on the platform to evaluate Medtronic's AccuRhythm AI in LINQ II loop recorders (20 percent less alert burden, 40 percent less episode burden), which is evidence about the device maker's algorithm. A published case study describes athenahealth integration at a health system and clinician training of about 15 minutes. No sensitivity is published for Auto-Triage, so the alert reduction says how much was removed and not what was removed with it.

General commitments are published and the secondary use of patient data is not governed in public. The privacy policy (PaceMate, LLC, updated June 2023) covers the website and services, says data is used to provide, maintain, improve and enhance the services and to generate anonymized, aggregate, de-identified data for purposes such as reports, that retention lasts as long as necessary, and that the company uses physical and electronic safeguards; it does not mention HIPAA.

The company also markets a research grade dataset combining device and electronic record data that it says outside parties use. Nothing states whether customers consent to their patients' data entering that dataset, the de-identification standard (hard to meet for longitudinal device records with serial numbers), who the users are, whether access is paid, or whether a clinic can decline and remain a customer, and nothing says whether data trains Auto-Triage. No safety engineering is published.

HIPAA compliance is stated and no business associate terms are published. The home page states HIPAA compliance alongside ISO/IEC 27001:2022, HITRUST r2 and SOC 2, and its footer links a notice of privacy practices as a PDF; the website privacy policy (June 2023) covers the website and services and does not mention HIPAA. The information security certifications answer who may access data and how it is protected, not what a business associate agreement commits to. No business associate agreement, retention position for patient data, or terms for data flowing to and from partners such as AliveCor's KardiaPro platform are published.

Certifications are named with type and version and can be checked with their issuers. The home page lists ISO/IEC 27001:2022 with its certificate number (ISMS-PA-052626), HITRUST r2 and SOC 2, and shows a penetration testing badge; the company announced the ISO certification in June 2026 from an accredited certification body.

An ISO certificate number can be verified with the issuing body, and HITRUST r2 certifications are listed in HITRUST's public directory, so a buyer can confirm two of the three without asking the company; the SOC 2 report itself is not published, and there is no trust portal, vulnerability disclosure policy or subprocessor list.

No device claim is made and no regulatory position is stated. PaceMate presents itself as a data management, workflow and billing platform that aggregates and displays data from already cleared implanted devices, which can sit outside device regulation.

The tension is Auto-Triage: software that filters which device alerts a clinician sees sits closer to the regulated boundary than software that only displays them, particularly where suppression has clinical consequences, and nothing published sets out the company's regulatory analysis, intended use or why the triage sits outside device classification. Whether the basis of a filtered alert remains open to the clinician is not described. The company's demonstrated regulatory competence is in payment rules: named procedure codes, 31 and 91 day billing cycles and maintained code updates.

Nothing is published on how Auto-Triage is governed or how it performs across patients: no model documentation, training data description, triage performance figure, subgroup analysis or bias statement. The Heart Rhythm study the company promotes (2024, de-identified data from 26 US centers on its platform) evaluated Medtronic's AccuRhythm AI embedded in LINQ II loop recorders, not PaceMate's own triage, reporting 20 percent less alert burden and 40 percent less episode burden.

The company markets a research grade dataset of device and record data and its clinical authors presented at a health equity session at HRS 2026, so the data and interest exist without a published breakdown. The exposure is specific: alert patterns differ by device type, age, comorbidity and how reliably a patient transmits, and patients with irregular connectivity, disproportionately older, poorer and rural, are those least likely to have a suppressed alert noticed.

A clinician signs what surfaces, and nothing stands behind what the triage suppresses or the billing automation submits. The platform's one click sign and next workflow keeps a clinician reviewing and signing each report that reaches them, which lets them challenge the platform's output before it enters the record. Two exposures remain unaddressed.

Auto-Triage decides which alerts a clinician sees, a filtered alert is reviewed by nobody, and only a reduction figure (61 percent less alert burden) is published, with no sensitivity or count of true events filtered, so a clinic cannot size its residual risk.

And Auto-Script automates remote monitoring billing documentation and missed charge recovery, where coding errors can expose a practice to payer recovery and, in the wrong circumstances, false claims liability; nothing states the human review between automated capture and a submitted claim. No indemnity, accuracy warranty or service level is published for either.

Bidirectional integrations with major record systems are named, with standards and production evidence. The home page lists Epic as bidirectional and SMART on FHIR verified, Oracle Health as bidirectional, and athenahealth as bidirectional through its FHIR API; device data is paired with record data in the platform and billing data synchronizes back.

A published case study describes the athenahealth integration at a health system as effectively plug and play, with the deployment later extended from implanted devices to ambulatory monitoring. On the device side the company cites 14 integrated manufacturers and Medtronic's SessionSync for in clinic wireless workflow, without listing the manufacturers.

The options and the location are stated, and the isolation and processing path are not. The platform is cloud based with no local installation, offered as self service, periodic support or a fully tailored service, which changes who touches patient data; the privacy policy (June 2023) states that the services are hosted in the United States, with possible transfers to other countries.

No hosting provider, region, tenancy model, subprocessor list or contract end terms are published, and nothing says where the research grade dataset the company markets is held or how it is separated from operational data.

No price is published, and the revenue side of the purchase is documented in detail. The company names the remote monitoring procedure codes it handles, tracks the 31 day and 91 day cycles that make a transmission billable, runs missed charge recovery, maintains annual code updates, synchronizes billing data with Epic, Oracle Health and athenahealth, and reports 75 percent average billing capture within 90 days of go live; it offers delivery as self service, periodic support or a fully tailored service.

That lets a device clinic estimate what it could recover from payers, which is the payer's side of the ledger rather than what the clinic pays PaceMate. No per patient or per clinic rate, platform fee, tier price, implementation cost, term or minimum is published.

Coverage is named broadly, with validation behind part of it. The company describes one platform for implantable cardiac devices, ambulatory event monitoring, heart failure monitoring and consumer ECGs (through AliveCor's KardiaPro integration), cites 14 integrated device manufacturers without naming them, and reports more than 2.3 million patients managed.

Named deployments (Ascension Jacksonville, Piedmont Healthcare, a health system that extended from implanted devices to ambulatory monitoring) stand behind the implanted device and ambulatory settings; the manufacturers supported at discrete data level are not listed, and nothing validates Auto-Triage across device types.