Can your Apple Watch really measure blood pressure?
What Apple's hypertension notifications detect, why they do not provide a blood-pressure reading, and where the feature's limits matter.
Healthcare Edition / Coverage
Health technology
Digital tools, medical technology and the evidence behind their use in care.
Start with your question
- What is recorded?
Identify the signals and recording conditions described for the feature.
- What is calculated?
Look for the statistic, algorithm or scoring method behind the display.
- What does it establish?
Read the validation evidence and limits alongside the number.
Supporting reading: Can your Apple Watch really measure blood pressure? · Apple Watch ECG, rhythm alerts and AF History: what differs? · Heart rate variability: what your wearable’s HRV reading means · What is VO₂ max, and how accurate is your smartwatch’s estimate?
Make sense of a wearable reading
Start with what the device reports, then explore how that number is calculated. These guides separate a notification, a measurement and an estimate, with the relevant device and evidence limits.
Explore 4 articles
Apple Watch ECG, rhythm alerts and AF History: what differs?
A short electrical recording, occasional background checks and a weekly estimate serve different purposes. Compare the inputs, outputs and intended users of three rhythm features.
Heart rate variability: what your wearable’s HRV reading means
HRV readings depend on how and when they are measured. Here is how personal baselines, overnight averages and device labels change their meaning.
What is VO₂ max, and how accurate is your smartwatch’s estimate?
A watch estimates VO₂ max without measuring your breathing. Learn what the number means, how Apple calculates it and what validation studies can establish.
A sleep score, an alert or a treatment?
A wearable rating, a possible sleep-apnoea notification and a structured insomnia programme have different purposes. These guides explain the documented inputs, intended use and evidence for each, with access and assessment limits kept separate.
Explore 3 articles
Your sleep score changed: what wearable ratings actually measure
Apple and Google turn sleep data into different ratings. Here is what feeds those numbers, why an update can change them, and what a score cannot establish.
Your Apple Watch flagged possible sleep apnoea: what does that mean?
The alert looks for repeated breathing-disturbance patterns over a month. Its observation window, intended users and report explain what it can contribute to a separate clinical assessment.
Sleepio and sleep trackers answer different questions
A digital CBT-I programme aims to treat insomnia; a tracker summarises sleep. NICE’s evidence and recommendation show why the distinction matters.
Look behind a home test result
A result has a method behind it. Explore what a validation listing establishes, what a hormone test detects and what a microbiome report can support. Each guide keeps its own scope and sources.
Explore 3 articles
Why a validated blood-pressure monitor still needs the right cuff
Exact-model validation and cuff fit answer different questions. Registry criteria and a crossover trial explain what to check behind a home reading.
Home menopause tests: what an FSH result can tell you
A urine strip can detect a hormone without diagnosing menopause. Here is why symptoms, timing and the kind of test change what a result means.
Gut microbiome tests: what can the report actually tell you?
A list of microbes, a comparative score and a useful clinical decision require different evidence. Testing research explains the gaps between them.
Understand how connected assistance works
Follow the feature beyond the hardware: who supplies an answer, how hearing assistance is fitted, and what happens between a fall alert and an attempted call. Model, setup and country limits stay with each article.
Explore 3 articles
The two kinds of help inside AI glasses for sight loss
Generated descriptions and live video calls to people offer different kinds of assistance. Here is how those routes differ on supported Meta glasses.
Hearing aids in earbuds and glasses: how the technology differs
AirPods and hearing glasses combine familiar hardware with amplification. Their fitting methods, software modes and evidence are different.
Your Apple Watch detected a fall. What happens next?
Movement, settings, a countdown and a working connection all matter between a fall alert and a call for help. Apple’s documentation explains the sequence.
Understand a health-technology study
Funding, test performance and use in a care service answer different questions. Follow these reports into their methods and limits, keeping the stage of each research project in view.
Explore 3 articles
ARPA-H commits up to $117.4 million to Delphi wearable biosensors
Four Delphi teams will develop wearable sensors for heart failure, inflammation and perimenopause, with funding tied to research milestones.
The search for Parkinson’s in a blood sample
An experimental electronic assay measures alpha-synuclein forms in blood-derived vesicles. Its small study shows why classification and future-disease prediction must stay separate.
Hospital at home: what Oxford’s wearable-patch study showed
A small Oxford study tested whether vital signs reached a clinical dashboard. Its findings distinguish data availability from accuracy and better patient outcomes.
Follow AI into clinical work
Explore what clinicians ask, how evidence is brought into an answer and what happens to documentation time. Each report separates a tool’s documented role from the outcomes its evidence can establish.
Explore 4 articles
What doctors ask OpenEvidence
A US study of 271 million queries describes how clinicians use a medical AI platform. Its question categories reveal information needs, while answer quality and patient benefit remain unmeasured.
Doximity Ask and UpToDate: what the planned integration adds
Doximity and Wolters Kluwer plan to place UpToDate content alongside Ask answers for clinicians at health systems using both products. That changes the information available in the workflow, while access and answer quality remain separate questions.
AI can write the notes. Do doctors get their time back?
A randomized crossover trial found differences in note-writing time without clear reductions in after-hours EHR work. The distinction matters when evaluating an ambient scribe.
Why health chatbots can stumble in conversation
A controlled study of UK adults shows why medical knowledge scores and useful conversations are different tests of a health chatbot.
Find your way through device oversight
Recognition, surveillance, public records and guidance priorities serve different purposes. These reports explain the relevant Great Britain, EU and US systems without treating their requirements as interchangeable.
Explore 4 articles
CE-marked medical devices: Great Britain’s plan for indefinite recognition
The government plans indefinite recognition for devices meeting EU MDR or IVDR rules. Legislative changes remain to come, while registration, UK representation and post-market duties continue.
Medical-device surveillance in Great Britain: PMSR, PSUR and vigilance explained
How MHRA guidance distinguishes routine PMSR and PSUR updates from serious-incident reporting in Great Britain, with report cycles, request deadlines and scope.
EUDAMED explained: modules, public records and rollout
What EUDAMED records, which modules became mandatory in May 2026, and what a public device or certificate entry can tell you.
FDA’s fiscal 2027 guidance priorities: AI devices and surgical robots
The FDA plans final guidance on AI device software and surgical robots, plus a draft on conversational AI for mental disorders. The priority list does not guarantee publication.
- The measuring system
The blood-pressure guide separates exact-model validation from cuff fit.
- The detected signal
The menopause guide separates urine FSH detection from diagnosis.
- The report’s meaning
The microbiome guide separates detection, reference comparisons and clinical utility.
Supporting reading: Why a validated blood-pressure monitor still needs the right cuff · Home menopause tests: what an FSH result can tell you · Gut microbiome tests: what can the report actually tell you?
Health technology
Retatrutide versus Ozempic: what the 80-week trial compared
Lilly reports greater weight and blood-sugar changes in adults with type 2 diabetes. The comparator, hypothetical treatment conditions and unpublished full results define what can be concluded.
Health technology
RadNet’s ONRAD deal is about who reads the scan
The proposed acquisition would add contracted radiologists and hospital staffing to RadNet’s imaging and software businesses. Those services play different roles in the reporting workflow.
Health technology
A small language model can label a radiology report without reading the scan
A single-centre study trained compact models to extract neck-fracture labels from written CT reports. Its results explain the difference between text extraction, image diagnosis and clinical deployment.
Health technology
A wearable forecast could time a movement reminder—but has not tested its benefit
Researchers used activity-tracker data from women with chronic pelvic pain to forecast sedentary periods. The analysis tested prediction and alert tradeoffs, rather than whether prompts reduce pain.
Health technology
Oncodetect’s new blood test looks for traces after cancer treatment
The tumour-informed test tracks DNA matched to a patient’s cancer. Its launch raises separate questions about detection limits, recurrence prediction and whether acting on a result improves care.
Health technology
Why an Ebola vaccine is being tested against Bundibugyo
The Tumaini trial has begun in DR Congo with Ervebo, a vaccine licensed for a different Ebola virus. The study is designed to establish evidence that existing approval cannot supply.
Health technology
Doctolib’s French consultation assistant: what happens to the notes?
Standalone and integrated versions take different routes into the medical record. French support documents explain where clinician review, copying and temporary working notes fit.
Health technology
What your iPhone’s Walking Steadiness alert measures
The Health app uses everyday walking data to describe gait-related steadiness over time. Its Low and Very Low labels serve a different purpose from an Apple Watch hard-fall alarm.
Health technology
Prostate MRI AI: what NICE’s draft still needs to establish
NICE proposes five tools for use alongside evidence generation. The consultation asks about accuracy, biopsies, pathway effects and different patient groups before routine adoption can be assessed.
Health technology
An arthritis AI decision aid was better at preferences than facts
A ten-case study and a separate 34-person evaluation ask different questions about an AI aid for juvenile arthritis: factual reliability, treatment reasoning and preparation for a clinical conversation.
Health technology
Speech AI can rate a recording without proving it improves speech
Whisper-ST1 classified severity in a small held-out speaker sample. Its technical results do not establish longitudinal monitoring, useful exercise selection or better rehabilitation outcomes.
Health technology
Encora X1: what the tremor wristband is cleared to do
A prescription wristband uses vibration to ease hand tremor during use. Its US indication, study comparisons and limits explain the claim.