Dr. Phillips described hypertension as a massive public health issue and emphasized that Apple wanted to “build this feature to work really well for people.” While the company used data from over 100,000 participants to develop the feature, they expect a million people to use it in the first year. “We clinically validated it with 2,229 adults over 30 days in a clinical validation study, and we’re really happy with the results.”

The notifications target users above age 22 who haven’t been previously diagnosed with hypertension and aren’t pregnant. After a one-time setup through the Watch or Health app, the feature runs automatically in the background. If you’ve already been diagnosed, the feature won’t activate, though you can still log blood pressure readings manually. The feature works with Apple Watch Series 9, 10, 11, Ultra 2, and Ultra 3, and generated reports can be shared via email, messages, or WhatsApp for medical consultations. If you receive a hypertension notification, Apple recommends logging your blood pressure for seven days using a third-party blood pressure cuff.

In clinical validation, Apple reports the feature achieved 41.2% sensitivity in correctly identifying people with hypertension and 92.3% specificity in identifying people without it—performance comparable to clinical BP cuffs, which typically show around 50% sensitivity and 90% specificity.

“When we look at hypertension, we see a disease that affects 1.3 billion people worldwide, and almost half are undiagnosed. The biggest impact we can make is helping that large undiagnosed population get an alert that nudges them toward diagnosis and blood pressure checks, preventing complications down the road,” Dr. Phillips said. He noted that hypertension is the leading modifiable risk factor for heart attacks, stroke, and kidney disease.

When asked whether the watch might produce false positives during stressful moments or when users are on certain medications, Dr. Phillips clarified: “Because hypertension is a long-term chronic disease, we designed our feature to detect that chronic condition. If you’re really stressed out or have just started medication—something that would raise your blood pressure only temporarily—that won’t trigger an alert.”

On the balance between sensitivity and specificity, he explained the priority was ensuring “we weren’t falsely alerting people that they have the disease when they don’t.” “While we aren’t catching everyone with hypertension, when we do send a notification, we’re highly confident it’s accurate. If you get a notification, it likely means you have hypertension. However, the absence of a notification doesn’t mean you don’t have it.”