Why the tonometer “undershoots” the upper pressure - and how to fix it without new devices

The classic measurement of pressure with a cuff and stethoscope (auscultatory method) systematically underestimates systolic pressure and overestimates diastolic pressure. A team of Cambridge engineers demonstrated the physical cause of this error in an experimental model and proposed simple calibration methods - up to changing the position of the hand - that can improve accuracy without replacing the devices. According to the authors and previous studies, due to the systematic underestimation of the upper pressure, up to 30% of cases of systolic hypertension may remain unrecognized. The study was published on August 12, 2025 in PNAS Nexus.

Background

• We measure pressure with a “cuff” for >120 years — but the standard remains invasion. There is a stable gap between non-invasive cuff methods (auscultatory by Korotkoff sounds and automatic oscillometric) and true intravascular pressure: in clinical comparisons, the cuff usually underestimates systolic and overestimates diastolic relative to intravascular recordings. This has been shown in recent reviews/meta-analyses and in works with a simultaneous arterial line.
• The physics of Korotkoff tones is more complex than it seems. The classics teach that as the cuff deflates, the vessel opens and then collapses — and we hear tones. But the detailed “mechanics” of the tone window and the factors that shift it have long remained the subject of hypotheses. Reviews have noted many influencing variables — from the shape of the artery and the rate of deflation to “pressure below the cuff” (in the forearm), which standard models rarely reproduce.
• Why is this critical for systolic hypertension? Diagnosis and treatment today are largely tied to systolic thresholds; if the upper pressure is systematically underestimated, then some people with isolated systolic hypertension (especially the elderly) remain underdiagnosed or undertreated. Hence the increased interest in the sources of systematic error in the method itself.
• Oscillometry is also “shaken” by algorithms. Automatic tonometers do not listen to tones, but analyze cuff oscillations and then recalculate them into SBP/DBP using proprietary (and closed) algorithms. This adds variability between models and does not eliminate the basic hydromechanics under the cuff. Therefore, even an ideally executed measurement technique does not always “fix” the systematic shift.
• Measurement technique still solves half the problem. Incorrect arm position/support, the wrong size cuff, talking during measurement, recent coffee/nicotine - all this can shift the numbers by many mmHg. AHA/ACC recommendations and European guidelines insist: arm at heart level and supported, cuff of the right size, 2-3 repetitions with an interval, rest for 3-5 minutes, legs uncrossed. Even a small thing like arm position in the JAMA IM study significantly changed the readings.
• What was missing from the current work? Although clinicians have long recognized that the cuff “loses” part of systole, there was no mechanistic explanation given the real-world “complete vessel closure + low pressure distal to the cuff” scenario: laboratory models typically used round tubes that did not fully collapse. The Cambridge study reproduces complete closure and shows how low “downstream” pressure delays arterial reopening by shifting the tone window—hence the systematic underestimation of SBP/overestimation of DBP.
• Why does a clinic need this: calibration without new devices. Understanding the role of "downstream" pressure gives ideas for protocol amendments (standardized hand position/maneuver before bleeding) and the potential for software adjustments in automatic devices - that is, a way to increase accuracy without a total change of the tonometer fleet.

What exactly did they do?

The scientists assembled a physical setup that reproduced the key conditions of the "cuff" measurement: compression of the "artery", cessation of flow below the cuff and its subsequent slow release. Unlike previous models with a round rubber tube, here they used flat-falling channels that, like a real artery under the cuff, completely close when pumped. This made it possible to test the effect of low pressure "downstream" (in the forearm) - a mode that occurs in a real arm when the brachial artery is compressed.

The main finding is the “delay in reopening”

When the cuff compresses the vessel, the pressure in the vessels below the cuff drops sharply and remains at a low "plateau". When the cuff is released, it is this pressure difference that causes the artery to remain closed longer than we expect - the "window" of the appearance of Korotkoff sounds (by which the upper/lower is counted) shifts, and the device/observer reacts later. The result is that the systolic pressure is underestimated, and the diastolic pressure is overestimated. The lower the "downstream" pressure, the greater the error. Previously, this mechanism was simply not reproduced in laboratory models, so the phenomenon of "floating systole" remained a mystery.

Why is this important?

• Hypertension is the #1 risk for premature death. If the upper pressure is consistently low, patients may not be diagnosed/treated. Reviews and clinical comparisons have previously documented discrepancies between cuff and invasive (intravascular) systole; this new work explains why.
• Solutions — without a total upgrade of the equipment. The authors show that accuracy can be increased protocol-wise — for example, by raising the arm in advance (creating a predictable “downstream” pressure) and then taking into account a predictable correction; in the future, devices can take into account age/BMI/tissue properties as a proxy for “downstream” pressure for individual correction.

What this changes now (for clinics and at home)

• For healthcare workers. In addition to standards for proper measurement (cuff selection based on arm circumference, “arm at heart level,” back supported, legs uncrossed, silence for 3–5 min before measurement, at least two repeat recordings), it is worth monitoring arm position and considering a uniform “lift-lower-measure” technique before deflation as a potential calibration. Official guidelines do not yet require this, but the work sets the direction for updating protocols and clinical trials.
• For people who measure at home. The "cheapest" increase in accuracy is the correct technique: a cuff of the right size, the arm at heart level and resting on the table, do not talk, sit quietly for 5 minutes, take 2-3 measurements with an interval of 1 minute and average. These steps themselves reduce the error much more than "upgrading" the gadget.

How does this fit into the race for new pressure technologies?

While searching for calibrations of the “classic”, alternative approaches are being developed in parallel – from optics (SCOS) to ultrasound (“resonance sono-manometry”) for continuous and cuff-less monitoring. But they will also face issues of validation and hydrostatic corrections. The new physics of the classical method is important already because the cuff will remain the main method in clinics and at home for a long time – and it can be made more accurate.

Limitations and the next step

The study provides a mechanistic explanation in a physical model and suggests protocol solutions, but now clinical trials are needed: to what extent the proposed techniques (for example, standardized hand position before measurement) correct the underestimation in real patients - at different ages, body types and with comorbidities. The Cambridge team is already looking for partners for such studies.

Source: Bassil K., Agarwal A. Underestimation of systolic pressure in cuff-based blood pressure measurement, PNAS Nexus 4(8): pgaf222, August 12, 2025. https://doi.org/10.1093/pnasnexus/pgaf222