The Week in Blood Pressure & Wearables — November 13-19, 2026

By Yair Lurie, Founder and CEO, Cardiowell. LinkedIn Newsletter edition, November 20, 2026.

This week's deep dive is about calibration. Three items landed in the same few days: a regulatory read of Nanowear's cleared cuffless garment, a smart ring called the "Best Blood Pressure Ring" by Forbes Vetted, and a review showing that how tightly a sensor is worn changes the optical signal. Each one comes back to the same requirement. The cuffless number still depends on a cuff reading.

Also this week: a meta-regression of 101 trials put a number on how much each mmHg of average systolic pressure moves control rates, a one-lot propranolol recall, a positive EU opinion for baxdrostat, home readings from a national remote monitoring program used to study blood pressure variability, and HIPAA changes that speed up patient access through apps.

Deep dive: What cuffless blood pressure devices still need from a cuff

Nanowear's SimpleSense-BP: cleared and calibrated every 21 days

Regulatory specialist Srividya Narayanan walked through Nanowear's SimpleSense-BP on November 16 in her newsletter, Reg-AI with Sri. Her headline, "Cuffless Does Not Mean Cuff-Free," sums up the piece. I went back to the primary sources.

SimpleSense-BP is a chest garment. It records two ECG leads, thoracic impedance for breathing, heart sounds, and activity and posture. Software combines those signals with age, gender, height, and weight to estimate systolic and diastolic pressure, per the FDA's 510(k) summary in its public device database. FDA cleared it on December 8, 2023.

The summary sets the boundaries:

  • Adults 22 and older, used "under the direction of a licensed medical professional," at rest with a stationary torso.

  • A clinician first sets the patient's hypertension class from an auscultatory cuff reading. A new cuff reading is "required every 21 days."

  • The comparison table lists the monitoring type as "Intermittent, Spot-checking." Readings are reviewed after the fact, with no alarms.

  • It isn't for use with a pacemaker, and readings taken during movement are rejected.

Nanowear's January 2024 announcement called it continuous blood pressure monitoring. FDA's summary describes spot checks.

The peer-reviewed study is in the Journal of Cardiovascular Translational Research (2025). It enrolled 303 adults: 185 in Pennsylvania and Georgia to train the model, and 118 in New York and New Jersey to test it. Compared with the reference readings, the mean error was -0.07 mmHg systolic (SD 6.71) and -0.58 mmHg diastolic (SD 5.89). The authors report that the device overestimates at lower pressures and underestimates at higher ones. They also state it "does not meet ISO 81060-3 criteria for continuous monitoring," and there was no overnight data. Nanowear funded the study, and four authors are Nanowear employees.

For engineers: ISO 81060-2 covers intermittent non-invasive monitors, the standard most home cuffs are validated against. ISO 81060-3 covers continuous monitors. The difference matters when a product is marketed as continuous.

Narayanan's main point is one I agree with. A near-zero average error can hide large errors for individual people. The standard deviation tells you more.

Circular Ring 2: a "best" label in a guide built around cuffs

Circular CEO Amaury Kosman posted on LinkedIn that Forbes Vetted named the Circular Ring 2 the "Best Blood Pressure Ring."

The label is real. It appears in Forbes Vetted's guide, The Best Blood Pressure Monitors: 6 Clinically Validated Machines To Track Your Vitals, dated October 12. The six top picks are validated cuffs. The ring appears under "Other Blood Pressure Monitors To Consider." Forbes says it "must be first calibrated with a traditional cuff-style blood pressure monitor," and that these wearables "aren't clinically validated."

Kosman was direct in the comments. He wrote that setup takes three cuff readings, then the app asks for a new one every month, or sooner if the resting baseline shifts. He also wrote, "It isn't a medical device, and it doesn't replace a cuff." Circular's blood pressure page says the ring "does not diagnose hypertension and is not a substitute for a cuff or for clinical advice." I found no FDA clearance or CE marking for its blood pressure feature.

The ring estimates blood pressure from pulse transit time. It uses its single-lead ECG and its optical (PPG) sensor together, timing the gap between the heartbeat and the pulse reaching the finger, per Pulse 2.0 when the feature launched in September. Circular's own page describes it more loosely, as reading the shape and timing of the pulse wave.

Holger Redtel, who develops calibration methods and tests wearables against ISO 81060, raised the sharpest technical point under the same post. Calibrating once, twice, or three times at the same resting pressure is still a single-point calibration. To know the size of a blood pressure change, you need calibration at two different pressure levels. He asked what reference data show the ring tracks both the direction and the size of changes in the same person. That question applies to most calibrated cuffless devices.

The fit of the sensor changes the signal

A systematic review published November 15 in npj Biosensing, from Khalifa University, Newcastle University and MIT, looked at contact pressure between the sensor and the skin across 29 studies of PPG, blood pressure estimation and related pulse sensing. Paul Sonnier shared it in his Digital Health group.

  • Too little pressure reduced optical stability and added motion artifacts.

  • Too much pressure compressed the small vessels near the surface, flattened systolic peaks and suppressed the dicrotic notch.

  • The best pressure depended on body site, sensor and outcome. The authors found no single target.

  • A cleaner waveform didn't by itself mean a lower blood pressure error.

  • Studies measured and reported contact pressure inconsistently, and the authors call for standardized, pressure-aware practices.

It's an early-access version that may still be edited. The suppressed dicrotic notch matters for any blood pressure model that uses the notch as a feature.

What I think

Every device in this deep dive leans on a cuff reading. Nanowear asks for one every 21 days. Circular asks for three at setup and one a month. The review adds a variable most products don't record: how tightly the sensor is worn when the calibration is taken, and every time after.

So the calibration reading anchors every number that follows. If that reading comes from an unvalidated cuff, with the arm unsupported or the patient mid-conversation, every estimate inherits the error. Last week we covered an AHA poster showing that common measurement mistakes moved office readings by 1 to 3 mmHg, up or down.

A clinic using a calibrated cuffless device still needs a validated home cuff, a calibration schedule, and someone responsible for making sure the reading happens and gets entered. That's a workflow cost the word "cuffless" hides. To be economically viable, someone has to own that step.

This is part of why we use Cardiowell’s ‘always-connected’ blood pressure monitor in the programs we support. Each reading goes over cellular to the care team with no pairing, so the reading—and any calibration that depends on it— arrives with a timestamp and a device behind it.

I'm optimistic about where this goes. Biozen's fingertip device, covered last week, shows that some designs can skip cuff calibration for spot checks. I hope the next generation of wearables logs contact pressure and calibration quality with every reading, the way good cuffs already flag arm movement.

What to watch

  • Whether Circular publishes validation data with its calibration protocol and reference method.

  • External validation of SimpleSense-BP, and any overnight data.

  • Whether wearable makers start reporting contact pressure, as the npj Biosensing authors recommend.

  • Whether any calibrated cuffless device publishes two-point calibration results, as Redtel asked.

Clinical & Research

Each mmHg of average systolic pressure moved control rates about 2 points

A meta-regression in Hypertension pooled 101 randomized trials: 65 individually randomized (57,201 participants) and 36 cluster randomized (28,154), per the journal. Hypertension Australia shared it on LinkedIn this week. A. Salam and colleagues at The George Institute led it. It found that every 1 mmHg drop in a group’s average systolic pressure came with about 2 more patients out of 100 reaching control (under 140/90).

For a clinic with 200 hypertension patients, that’s about 4 more controlled patients per mmHg. Lowering the panel’s average from 140 to 130 took control from about 45% to about 65%. Reaching 80% took an average of about 123 to 125. The average has to sit well below the target, because many patients hover just above it.

It’s about a group of patients, not one person. If you take the average systolic reading across a panel of patients and lower it by 1 mmHg, about 2 more patients out of every 100 end up with their blood pressure under control. “Points” means percentage points: control goes from, say, 45% to 47%.

Control was defined as under 140/90. Each 1 mmHg lower mean follow-up systolic pressure was linked to 2.1 percentage points higher control in individually randomized trials and 2.4 points in cluster trials. Mean systolic pressure explained about three-quarters of the variation between studies. In individual trials, predicted control was about 45% at a mean of 140, 65% at 130, and 85% at 120. Reaching about 80% control took a mean near 123 to 125.

The limits: it pools averages from trial arms, without patient-level data. The individual trials were mostly short drug trials; the cluster trials varied a lot, and the analysis covers only the 140/90 threshold. It shows an association across trials. I couldn't open the full paper; the figures come from a published summary and the journal listing.

For clinics, it's a useful planning number. If a practice wants most patients controlled, the average reading has to sit well below the threshold.

Home readings show more blood pressure variability in neurodegenerative disease

A University of Miami study in the Journal of Clinical Hypertension used de-identified data from a national remote patient monitoring program with 40,278 participants, per the Miller School. Patients needed readings on at least three days a week for at least eight weeks. 49 people with Alzheimer's disease, dementia, or Parkinson's disease were matched to 147 controls.

Day-to-day systolic variability was 15.5% higher in the disease group, while mean systolic was about 133 in both. The case group is small, diagnoses came from billing codes, and medications weren't accounted for. The authors say variability isn't yet a clinical marker. The program and device weren't named.

The interesting part is the data source. Frequent home readings let researchers look at variability, which an average hides.

Heart failure telemonitoring at regional scale in Austria

HerzMobil Steiermark enrolled patients after a hospital stay for acute heart failure at nine centers from 2019 to 2025, per News-Medical, reporting on a presentation at the ESC Digital & AI Summit 2026. Patients sent daily weight, blood pressure, and heart rate data from a connected scale and blood pressure monitor, and nurses and physicians reviewed it. The quality-of-life score (KCCQ) rose from 62.6 to 76.9 at three months. 90% used the devices independently.

It's a conference presentation with no control group and no readmission or mortality data. The useful finding is operational: patients used the devices, and a nurse-and-physician team reviewed them across a whole region.

Policy & Regulation

One lot of propranolol recalled for possible metal contamination

Breckenridge Pharmaceutical recalled one lot of propranolol hydrochloride extended-release capsules, 60 mg, on November 16. The details:

  • Lot PK25110035, NDC 51991-817-01, 100-count bottles, expiring May 2029, made by Adare Pharma Solutions.

  • Shipped to wholesalers and pharmacies from October 26 to November 4, 2026.

  • The lot shared intermediate material with a separate, unreleased lot where metal was found. The risk is cuts to the mouth, esophagus, stomach or intestine.

  • The company knows of no related adverse events. Patients should stop taking capsules from that lot and talk with their doctor.

Cardiovascular Business notes FDA hadn't assigned a recall class. I didn't find it on FDA's recall list as of November 18. This is separate from the Macleods olmesartan/amlodipine/hydrochlorothiazide recall covered November 6. Propranolol is also used for arrhythmia and migraine, so not every patient on it is being treated for blood pressure.

Baxdrostat gets a positive opinion in the EU

The EMA's medicines committee recommended approval of Baxfendy (baxdrostat) on November 12 for adults whose blood pressure isn't controlled on at least two medicines, per AstraZeneca. The European Commission still has to decide. Baxdrostat is an aldosterone synthase inhibitor, the same class as lorundrostat, covered last week. FDA approved it in the US in May.

The company cites two trials. In BaxHTN (796 people), 2 mg lowered seated systolic pressure 9.8 mmHg more than placebo at 12 weeks. In Bax24 (218 people with resistant hypertension), 24-hour ambulatory systolic fell 14.0 mmHg more than placebo. The release doesn't give potassium figures. As with lorundrostat, a practice prescribing it will want potassium checks next to the home readings.

HIPAA changes speed up patient access through apps

HHS finalized updates to the HIPAA Privacy Rule, per MobiHealthNews and Fierce Healthcare:

  • Covered entities must respond to records requests within 15 calendar days, down from 30, with one extension of up to 30 days.

  • Patients can get their information through personal health apps when the covered entity is technically able.

  • More copies must be free, with fees disclosed in advance.

  • The rule takes effect 60 days after Federal Register publication, with compliance 180 days after that.

For remote monitoring, faster app-based access means a patient's readings and clinic notes can move to the tools they use. I didn't confirm the Federal Register date.

CMS: still no confirmed final RPM rates

The CY 2027 Physician Fee Schedule final rule addenda on CMS's page are dated November 5. I found no published per-code final rates for the remote monitoring codes from CMS or the trade press this week, and no CMS guidance defining "direct employee," which CMS has promised before the requirement starts January 1, 2029. One RPM vendor, RPM Healthcare, says in a release that proposed cuts to clinical labor for 99457 and 99458 weren't finalized. That's the company's reading, and I haven't checked it against the rule text. The general understanding is that the code for supplying the device and transmitting the results at least twice per month is reduced by 20%. Overall reimbursement for the solution will be about what it was in 2025.

EU device rules and AI

A comment in npj Digital Medicine reviews the European Commission's proposed reform of the Medical Device Regulation. It would move medical devices out from under direct application of the AI Act's high-risk chapter, allow more non-peer-reviewed and in-silico evidence, and add an "innovative pathway," regulatory sandboxes, and broader change-control plans. The authors say it improves things "only partially," and continuous-learning AI remains unaddressed. One author works at Philips. Adoption is expected in 2027-28. This affects how an AI blood pressure algorithm gets classified and validated in Europe.

Devices & Wearables

Whoop at HLTH: Medicare, GLP-1s and older members

At HLTH 2026 in Las Vegas, Whoop described a bigger health care role, per Fierce Healthcare. It takes part in CMS's ACCESS model for chronic care, it's part of Novo Nordisk's NovoCare direct-to-consumer program so GLP-1 patients can add Whoop data, and members 65 and older are its most engaged group. It offers clinician-reviewed lab tests and video visits, and has an FDA-cleared ECG feature for atrial fibrillation signs.

Blood pressure came up as something older members care about and a possible ACCESS outcome. Whoop's own Blood Pressure Insights is a wellness feature. FDA closed the July 2025 warning letter over it in June 2026, per MassDevice. For a hypertension outcome in a Medicare model, I'd expect validated cuff readings to be the measure.

Apple and OpenEvidence test AI health answers in up to 200,000 people

Apple and OpenEvidence launched a study app at HLTH, per STAT (paywalled) and Chief Healthcare Executive. It's open to up to 200,000 US adults 21 and older with an iPhone 15 Pro or later. OpenEvidence's AI answers health questions from peer-reviewed sources, and users can choose to connect personal data such as vitals or lab results. The study looks at trust, understanding and preparation for doctor visits, and Apple says the data won't train models.

If people bring blood pressure readings to an AI chat, the quality of those readings matters as much as the answer.

AliveCor brings its ECG AI to other watches

AliveCor and PulseAI will offer PulseAI's FDA-cleared ECG interpretation on Apple Watch, Galaxy Watch, Fitbit Sense and Withings ScanWatch through a KardiaCare subscription in the first half of 2027, per the companies' release. On Apple Watch it adds beat-by-beat PVC detection. The companies cite a 296-person study in Heart Rhythm with fewer inconclusive readings than Apple's own algorithm. Cleared third-party software on consumer hardware is a model blood pressure could follow.

Noom launches its own band

Noom will launch Noom Band in early 2027 for employer and health plan clients, per MedCity News. Built on Sensor Bio hardware, it tracks activity, sleep, heart rate, HRV, and SpO2. It doesn't measure blood pressure. Noom's own figures, not peer-reviewed, say GLP-1 members with a connected wearable were 34% more likely to lose 20 pounds by month four.

Also worth a glance

  • Nocturnal hypertension depends on the threshold. In 634 people with type 2 diabetes on 24-hour ambulatory monitoring, nocturnal hypertension was 62.5% under ESH/ISH thresholds and 83.3% under ACC/AHA, per EMJ. More than half of isolated cases changed category between systems.

  • Salt substitutes. A JAMA Medical News piece covers potassium-enriched salt substitutes, which fewer than 1 in 10 US adults use. It reports an FDA final rule letting food makers use them starting in December. They aren't advised for people with kidney disease or on potassium-raising drugs. I haven't checked the FDA rule itself.

  • HHS on AI doctors. HHS released a white paper on what it calls "Super Intelligence" doctors, covering state AI regulatory sandboxes and incentives for states and pilot participants to coordinate with federal agencies, per Second Opinion (paywalled). It isn't regulation. I haven't read the HHS document itself.

  • Huawei Watch D3 in Indonesia. The watch launched in Indonesia on November 19, per Palpos. It measures with an inflatable airbag in the strap, the oscillometric method, and offers 24-hour monitoring.

  • A troponin patch, in mice. McMaster and Waterloo researchers built a microneedle patch that tracks cardiac troponin in skin fluid over several days, per Talker News, in Science Advances. Mouse models only, with no accuracy figures in the coverage.

  • Pharmacies in Medicaid care. UnitedHealthcare is working with CPESN USA community pharmacies for Medicaid and dual-eligible members in seven states, with medication management and point-of-care testing, per Becker's. Blood pressure isn't named, but a pharmacy is a natural place for a check.

  • A wearable standard from CTA. The Consumer Technology Association plans to announce a Wearable Performance and Trust standard at CES 2027 in January, per its release.

  • What patients say about virtual care. Bridge, a virtual care company, surveyed 2,016 US adults in August. 91% of users were at least somewhat satisfied, 64% said a visit fully resolved their issue, and 56% got virtual care through their own provider, per the report. It's unweighted and vendor-run.

  • Veterans' tele-prescribing. The House passed a 62-bill veterans package on November 11 that extends VA clinicians' ability to prescribe controlled substances by telemedicine, per Rep. Womack and Stars and Stripes. The Senate hasn't acted.

  • RPM consolidation. Vivo Care, a remote monitoring and care management vendor, combined with Onpoint Healthcare Partners, per PR Newswire. Terms weren't disclosed.

The thread this week

This week was about anchors.

The meta-regression says control follows the average reading. The Miami study shows that frequent home readings reveal what an average hides. And every cuffless device in the deep dive still sets its numbers from a cuff reading, taken at rest, at one fit, at one point in time.

Wearables are getting better, and some are learning to skip the cuff. Until most of them do, the cuff reading behind them deserves the same care as the reading a clinic would titrate from.

This edition is also on the Cardiowell blog: [ARTICLE URL — cardiowell.com, not yet published].

If you'd like this in your inbox each week, subscribe to The Week in Blood Pressure & Wearables.

References

  1. Cuffless Does Not Mean Cuff-Free: What Did Nanowear Actually Build?. Srividya Narayanan on Substack (Reg-AI with Sri).

  2. 510(k) Summary, SimpleSense-BP (Nanowear Inc.). US Food and Drug Administration, 510(k) premarket notification database (link omitted).

  3. Nanowear Announces FDA 510(k) Clearance for AI-enabled Continuous Blood Pressure Monitoring and Hypertension Diagnostic Management: SimpleSense-BP. Business Wire.

  4. Development and Internal Validation of an AI-Enabled Cuff-less, Non-invasive Continuous Blood Pressure Monitor Across All Classes of Hypertension. Journal of Cardiovascular Translational Research (Lopez-Jimenez et al., 2025).

  5. Really happy to share this one: Forbes Vetted named Circular Ring 2 the Best Blood Pressure Ring!. Amaury Kosman on LinkedIn (comments include Holger Redtel's calibration question).

  6. The Best Blood Pressure Monitors: 6 Clinically Validated Machines To Track Your Vitals. Forbes Vetted.

  7. Circular Ring 2 blood pressure page (headline not confirmed). Circular.

  8. Sensor-skin contact pressure as a hidden confounder in wearable photoplethysmography biosensing. npj Biosensing (Elgendi et al., 2026).

  9. Could the fit of your smartwatch be skewing its blood-pressure reading? (headline from the post's opening, not confirmed). Paul Sonnier on LinkedIn (Digital Health group).

  10. Association Between Follow-Up Mean Blood Pressure and Blood Pressure Control: A Meta-Regression Analysis of 101 Randomized Trials. Hypertension (Salam et al., 2026). Full text not opened.

  11. September 2026 Newsletter. Hypertension Australia.

  12. Study Finds Greater Blood Pressure Variability in People With Neurodegenerative Disease. University of Miami Miller School of Medicine (InventUM).

  13. Telemedicine-based management program improves quality of life for heart failure patients. News-Medical.

  14. Breckenridge Pharmaceutical, Inc. Issues Voluntary Nationwide Consumer-Level Recall of One Lot of Propranolol Hydrochloride Extended-Release Capsules, USP, 60 mg, Due to Potential Metal Contamination. Breckenridge Pharmaceutical.

  15. Blood pressure drug recalled due to risk of patient harm. Cardiovascular Business.

  16. Baxfendy (baxdrostat) recommended for approval in the EU for the treatment of adults with hypertension. AstraZeneca.

  17. Baxfendy approved in the US as the first and only aldosterone synthase inhibitor treatment for adults with hypertension. AstraZeneca.

  18. HHS updates HIPAA for patient health app access. MobiHealthNews.

  19. HHS finalizes updates to HIPAA privacy rule to ease patient access to health information. Fierce Healthcare.

  20. CMS-1848-F (CY 2027 Physician Fee Schedule final rule). Centers for Medicare & Medicaid Services.

  21. RPM Healthcare Welcomes CMS Decision Preserving Physician Practices' Remote Monitoring Partnerships. EIN Presswire.

  22. The proposed reform of the Medical Device Regulation and its implications for AI-enabled medical devices in the EU. npj Digital Medicine.

  23. HLTH 2026: How Whoop aims to play a bigger role in healthcare. Fierce Healthcare.

  24. FDA closes warning letter over Whoop blood pressure insights. MassDevice.

  25. Apple and OpenEvidence launch experimental health chat app. STAT (paywalled).

  26. Apple plans to study how consumers use AI for health needs: Questions and answers. Chief Healthcare Executive.

  27. AliveCor and PulseAI Partner to Bring Advanced Cardiac AI Determinations and Clinical Care Platform to Major Consumer Wearables (headline may be truncated). EP Lab Digest (HMP Global).

  28. Noom Introduces Wearable to Support Obesity Care Programs. MedCity News.

  29. Blood Pressure Thresholds Reshape Nocturnal Hypertension Diagnosis. EMJ.

  30. In an Effort to Reduce Hypertension, Salt Substitutes Come to the Table. JAMA.

  31. Huawei Watch D3 Meluncur 19 November, Bisa Pantau Tekanan Darah 24 Jam. Palpos.

  32. New sensor can detect 'silent' heart damage after surgery. Talker News via NBC Right Now.

  33. UnitedHealthcare, pharmacy network forge collaboration for Medicaid members. Becker's Payer Issues.

  34. CTA Previews Health Innovation at CES 2027, Announces New Wearable Technology White Paper. PR Newswire (Consumer Technology Association).

  35. The State of Virtual Care 2026. Bridge.

  36. Womack Votes to Support Combat Veterans and Improve Veteran Access to Telemedicine Services. Office of Rep. Steve Womack.

  37. House passes sweeping veterans bill (headline approximated from the URL; not confirmed). Stars and Stripes.

  38. Vivo Care and Onpoint Healthcare Partners Combine Around Provider-Led Remote Care. PR Newswire.

  39. Circular Adds Blood Pressure Monitoring To Circular Ring 2 Using ECG And PPG Sensors. Pulse 2.0.

  40. HHS to wrangle the state AI sandboxes. Second Opinion (Ruth Reader; paywalled).

Next
Next

The Week in Blood Pressure & Wearables — November 6-12, 2026