US, 23, Jan 2026 – The U.S. Food and Drug Administration (FDA) has officially released a new draft guidance titled “Cuffless Non-invasive Blood Pressure Measuring Devices – Clinical Performance Testing and Evaluation.“ This document provides the industry with long-awaited recommendations on how to validate the accuracy and safety of cuffless technology ranging from smartwatches to continuous monitoring sensors. For manufacturers currently developing or marketing these devices under Product Code DXN, this guidance represents a critical shift toward standardized clinical studies.
OVERVIEW OF THE NEW GUIDANCE FOR CUFFLESS BLOOD PRESSURE DEVICES
The scope of this guidance is specific to cuffless non invasive blood pressure (NIBP) systems. It clarifies that while traditional cuff-based devices and neonatal ICU monitors remain outside this specific scope, any cuffless device that derives systolic, diastolic, or mean blood pressure through sensors is now subject to these refined clinical performance expectations.
KEY TAKEAWAYS FOR MANUFACTURERS
1. Transparency in “Black Box” Technology
Manufacturers must now provide a deep dive into the technological principles of their devices. This includes:
- Sensor Identification: Clear descriptions of sensor types, such as photoplethysmogram (PPG) or electrocardiogram (ECG).
- Derived Parameters: How the device uses data like
- (PAT) or Pulse Wave Transit Time (PWTT).
- Algorithm Integrity: Documentation of AI/ML training datasets and signal quality checks used to filter “noisy” data.
- Calibration Needs: If a device requires a traditional cuff for “initialization,” manufacturers must specify how often this re-initialization must occur.
2. The Clinical Testing “Triple Threat”
The FDA recommends that clinical studies data be submitted from three distinct types of tests, evaluated independently.
- Static Test: Demonstrates accuracy immediately after the device is set up.
- Stability Test: Proves the device remains accurate over its entire “labeled window” of use (e.g., if a device claims to be accurate for 28 days, it must be tested at various points in that window).
- Change Test: Validates that the device can detect real-world blood pressure swings, specifically changes of at least 15 mmHg (systolic) and 10 mmHg (diastolic).
3. Rigorous Diversity Requirements
To combat sensor bias, the FDA has set high bars for participant demographics:
- Minimum Sample Size: At least 85 participants (ages 21–50) for standard adult use, with a minimum of 30% male and 30% female representation.
- Skin Tone Inclusivity: The agency recommends that at least 25% of subjects fall into each category of the Monk Skin Tone scale (1-4, 5-7, and 8-10) to ensure sensors work across all pigmentations.
- Special Populations: If the device is intended for seniors (over 50), children, or pregnant women, additional cohorts of 85 subjects per group are typically recommended.
ACTION PLAN FOR FOOD AND DRUG ADMINISTRATION INDUSTRY
The FDA “strongly recommends” that manufacturers utilize the Q-Submission Program before starting clinical trials. This allows for early feedback on study protocols, which is vital given the novelty of AI/ML techniques in this space. Acceptance Criteria: To meet FDA standards, devices should generally demonstrate a mean error ≤ 5.0 mmHg and a standard deviation < 8.0 mmHg across all tests.
TIMELINE & COMMENTS
This is a Draft Guidance and is not yet for implementation. The FDA is inviting stakeholders to submit comments and suggestions within 60 days of its publication in the Federal Register.