The FDA is warning the public to avoid using unreviewed devices for checking vital signs—including those for monitoring infants, as well as checking blood pressure at home—due to the risk of unverified and potentially inaccurate readings.
The pair of agency alerts comes as major companies begin to stake their first claims in this regulatory space—including just last week, with Apple obtaining an FDA clearance to add hypertension notifications to its Apple Watch. And in recent years, developers such as Owlet and Masimo have sought agency green lights for sensor-equipped baby booties, designed to keep an eye on a newborn’s blood oxygen levels.
According to the FDA, many devices sold over-the-counter and online today that claim to measure blood pressure, including some smartwatches and smart rings, have not been thoroughly tested.
They may deliver inaccurate results while also prompting the user to seek emergency medical care or change their medications, or they may miss the early signs of heart disease entirely—both of which can lead to harm, the agency said. The FDA said these types of devices require an official authorization to be marketed in the U.S., and that it is “taking steps to address” unlawfully sold products.
The agency said it has cleared multiple models of home blood pressure monitors, marketed under different brand names from dozens of companies, and urged users to search its database of authorized products.
There you’ll find the FDA’s summary of Apple’s latest clearance, which the tech company said will help it identify at least 1 million undiagnosed cases of high blood pressure within its first year.
Apple first submitted the artificial intelligence-based software for review in February. Using the same sensors the Apple Watch uses to track a person’s pulse rate, it now monitors how blood vessels react to the amount of force behind each beat. By calculating a 28-day average, it aims to alert the user to potentially chronic hypertension, the main risk factor for cardiovascular diseases such as heart attacks and stroke.
According to the document, its 510(k) clearance was tied to a de novo greenlight previously issued to a quite different piece of software: an algorithm for detecting hypertrophic cardiomyopathy put forward by Viz.ai.
The Viz HCM module, developed with backing from drugmaker Bristol Myers Squibb, scans 12-lead hospital ECG recordings to uncover the signs of thickened cardiac tissue in adults, tied to subtle changes in how electrical signals travel through the heart. It aims to offer an easier diagnostic pathway to the often hidden condition, which has few obvious symptoms and typically requires an ultrasound exam.
Though rare, HCM is one of the leading causes of sudden cardiac death for people under 35, and BMS has been looking to develop its first-in-class Camzyos (mavacamten) therapy into a blockbuster with an estimated potential of $4 billion in peak sales.
According to the FDA, the differences between those applications—including separate indications, settings, users and sensor technologies, with the Apple Watch chasing the much larger population of people with high blood pressure using an over-the-counter consumer wearable—did not raise new questions of safety or effectiveness. In its clearance, the agency said the Apple Watch software had been properly evaluated and validated through non-clinical and clinical testing.
Its development included a base machine learning model trained on Apple Watch data collected from more than 86,000 participants. A second model then constructed the hypertension algorithm by adding home blood pressure reference measurements from nearly 10,000 people.
A clinical study of about 1,860 participants went on to meet its primary endpoints, delivering a false-positive rate of 7.7% with a specificity of 92.3%, as well as a false-negative rate of 58.8% with an overall sensitivity of 41.2%. According to the FDA, that resulted in a positive predictive value of 70.9% when assuming that hypertension affects about one-third of the U.S. population.
Apple’s submission also included a predetermined change control plan, which allows the company to make basic updates and retrain its algorithm with new data without having to fully undergo a new round of FDA review.
Meanwhile, the FDA issued similar warnings in unauthorized baby monitors designed to record heart rates, oxygen saturation, body temperature and breathing.
The agency cautioned that these devices have not been evaluated for the ability to detect asthma or sleep apnea, as well as sudden infant death syndrome or sudden unexpected infant death, known as SIDS and SUID, respectively.
The FDA also said infant monitors are not a replacement for adult supervision or safe sleep practices. These include devices worn by the infant, as well as sensors placed under a mattress or wall-mounted cameras.
“The FDA is not aware of any clinical or scientific evidence that shows that currently available baby products prevent or reduce the chance of SIDS,” the agency said in its notice. “Be aware that any product marketed with claims to prevent or reduce the chance of SIDS or SUID has not been cleared or approved for that use by the FDA.”
In late 2021, Owlet was forced to pull its infant and toddler-monitoring Smart Sock from the market after a warning from the FDA, saying it required an agency green light. Nearly two years later, the company obtained the agency’s blessing for its updated BabySat system, as well as an over-the-counter clearance for the Dream Sock’s pulse oximetry features. Masimo, meanwhile, received a go-ahead in May 2024 for its Stork monitoring system.
The FDA has also granted a de novo clearance to the Snoo motorized bassinet from Happiest Baby, as its first device for encouraging proper baby sleep posture. Versions of the Snoo have been employed in hospital nurseries and NICUs; however, it has not been tested or cleared to reduce the risk of SIDS or SUID.