Hypertension Notification Feature (HTNF)

K250507

Apple Inc. · cleared 2025-09-11 · product code SFR · Cardiovascular

Premarket evidence — what FDA accepted

Device typesamd
source quote (p.5)
“The Hypertension Notification Feature (HTNF) is a software-only mobile medical application that analyzes photoplethysmography (PPG) data opportunistically collected by Apple Watch to identify patterns that are suggestive of hypertension and provides a notification to the user.”
Algorithmmachine learning (ML)-based algorithm comprising a deep-learning (DL) model and a Linear model
source quote (p.11)
“HTNF incorporated a machine learning (ML)-based algorithm to identify key PPG patterns that are suggestive of hypertension. The ML-based algorithm comprises of two key models: A deep-learning (DL) model developed using a self-supervised learning method based on a large-scale unlabeled data to extract generalizable characteristics of the PPG input signals. The unlabeled data included Apple Watch sensor data collected over 86,000 participants. A Linear model trained on top of the DL model to provide specific hypertension classifications (hypertensive vs. non hypertensive).”
Adaptive (vs locked)No
source quote (p.15)
“The PCCP does not include provisions for implementation of adaptive algorithms that will continuously learn in the field. All algorithm modifications will be trained, tuned, and locked prior to release of the software to the field.”
PCCPYes
source quote (p.2)
“FDA's substantial equivalence determination also included the review and clearance of your Predetermined Change Control Plan (PCCP).”
Cybersecurity addressedYes
source quote (p.11)
“Apple's approach to cybersecurity aligns with FDA's 2023 Guidance, "Cybersecurity in Medical Devices: Quality System Considerations and Content of Premarket Submissions." The device also conforms to the cybersecurity requirements identified in Section 524B to the FD&C Act.”

Validation studies (2)

Retrospective clinical

n=1,863 patients

endpoints: subject level sensitivity in identifying hypertension status; subject level specificity in identifying hypertension status; detecting possible hypertension over a 30-day period

Retrospective clinical

n=187 patients

endpoints: long-term specificity for non-hypertensives

Reported performance (7 observations)

sensitivity41.2CI [37.2, 45.3]
source quote (p.10)
“Sensitivity: 41.2% (95% CI [37.2, 45.3])”
specificity92.3CI [90.6, 93.7]
source quote (p.10)
“Specificity: 92.3% (95% CI [90.6, 93.7])”
ppvas written: “PPV (prevalence of 31.4%)”70.9CI [65.7, 75.7]
source quote (p.10)
“PPV (prevalence of 31.4%): 70.9% (95% CI [65.7, 75.7])”
sensitivityas written: “Sensitivity for stage 2 hypertension”53.7CI [47.7, 59.7]
source quote (p.12)
“The sensitivity for stage 2 hypertension was 53.7% (95% CI [47.7, 59.7])”
specificityas written: “Specificity for normotensive”95.3CI [93.7, 96.5]
source quote (p.12)
“the specificity for normotensive was 95.3% (95% CI [93.7, 96.5]).”
specificityas written: “Long-term specificity for non-hypertensives”86.4CI [80.2%, 92.5%]
source quote (p.14)
“the long-term specificity for non-hypertensives (N=187) remained high at 86.4% (95% CI [80.2%, 92.5%])”
specificityas written: “Long-term specificity for subset of non-hypertensives with normal blood pressure”92.5
source quote (p.14)
“as did the specificity for the subset of non-hypertensives with normal blood pressure (N=121) which was 92.5%”

Each value carries its own analysis unit and task — never compare or pool across devices. Source: 510(k) summary PDF.

Predicate network

Postmarket — what happened after clearance

0
recalls in product code, 24mo
0
MAUDE reports in code, 12mo
—
vs code's own 3-yr baseline
0
drift signals on this device

Recall and MAUDE counts are product-code-level (reports aren't reliably attributable to one device); a recall is shown as device-attributed only when the recall record itself lists this clearance number. Signals are descriptive observables with sources — never a judgment that the device is unsafe or drifting. Snapshot 2026-07-08.

Reimbursement — how devices like this got paid

Not yet tracked — no payment pathway indexed for this clearance (the reimbursement corpus is a growing seed set).

Applicable FDA guidance — what the submission is measured against

FDA guidance documents and guiding principles applicable to 510(k) AI/ML devices in the Cardiovascular panel. A curated reference index, not legal or regulatory advice — each item states its own status, and a draft is never binding.

Applicability is derived from the device's FDA advisory panel and pathway — cross-cutting guidances apply to every AI/ML device; panel-specific ones are flagged. Titles, dates, and links verified against fda.gov as of July 2026.

Constat Precedent · public FDA/CMS data · descriptive decision-support, not regulatory or reimbursement advice. Share this page: constat.dev/precedent/device/K250507