Chapter 2: Measuring and monitoring blood pressure

Blood-pressure measuring devices

Validated cuff-based devices

Accurate blood-pressure measurement is the cornerstone of diagnosing and managing hypertension. Direct intra-arterial measurement is the only method that provides the true arterial pressure, but its invasive nature prevents widespread clinical use. Non-invasive methods that can approximate intra-arterial pulsatile pressure have existed for many years, but their use has largely remained in research, except for selected diagnostic applications such as autonomic failure or orthostatic hypotension.

In clinical practice, blood pressure is therefore measured almost entirely by indirect, non-invasive methods. These methods use an inflatable cuff to occlude the brachial artery, assessment of the radial pulse, or a stethoscope to detect Korotkoff sounds. With theoretical and practical refinements, they remain almost universally used, and their adoption in clinical trials established the foundation for diagnostic and treatment recommendations.

Manual auscultatory measurement with a mercury or non-mercury sphygmomanometer remains the reference method for validation studies of new blood-pressure technologies. Mercury devices, however, have been removed from clinical use because of environmental toxicity. Good-quality hybrid auscultatory devices—with an LED or LCD display or a digital countdown—and shock-resistant aneroid devices can be used, but they remain vulnerable to observer-related errors such as terminal-digit preference, observer bias and preconceived judgement.

Automated cuff-based measurement, usually oscillometric, was developed to reduce the observer errors of manual auscultation. It is now the usual method for ambulatory blood-pressure monitoring (ABPM) and home blood-pressure monitoring (HBPM), and is also preferred for office blood-pressure measurement (OBPM). If an upper-arm cuff cannot be used, a validated electronic wrist-cuff device may be considered.

Cuffless blood-pressure devices

All cuff-based methods have limitations. They provide snapshots of blood pressure in relatively static conditions and do not fully capture its dynamic nature—its variation in response to daily activities and challenges. Errors caused by an inappropriate cuff size, shape or position are common, and limb compression during cuff inflation can be uncomfortable, particularly at work or during sleep.

New cuffless devices use sensors, signal processing, machine learning and other technologies embedded in wearable devices, smartphones or pocket devices. They may improve awareness and support future diagnosis and treatment, but important problems must be resolved before they can be recommended for routine clinical use. The standard AAMI/ESH/ISO validation protocols used for cuff devices are not sufficient for cuffless technologies; their accuracy remains incompletely established and there is no universally agreed performance standard.

Most cuffless devices also require periodic calibration against a standard upper-arm cuff and some require demographic information such as age and sex. In practice, many do not directly measure blood pressure. They track changes relative to a calibration measurement or attempt to predict pressure using demographic data and machine-learning algorithms. For these reasons, cuffless devices should not currently be used to diagnose or manage hypertension in clinical practice.

Standard office blood-pressure measurement (OBPM)

Conventional or standard OBPM is the most extensively studied method for evaluating blood pressure. It underpins the diagnosis and classification of hypertension, the role of blood pressure as a cardiovascular risk factor, evidence for the protective effects of antihypertensive treatment, and the definition of treatment thresholds and targets. Despite its limitations and the increasing use of out-of-office measurements, OBPM remains the most widely used approach for diagnosing and managing hypertension.

Although it has been used for a long time, OBPM is not sufficiently standardised in many settings. This produces inaccurate estimates, often overestimates blood pressure, and may lead to overdiagnosis and overtreatment. A standardised OBPM method should therefore control the measurement environment, patient position, device, measurement schedule and interpretation.

Measure blood pressure on a bare arm. Cuff size is critical and should be chosen according to the patient’s arm circumference. A cuff that is too small overestimates pressure, while one that is too large underestimates it. For manual auscultatory devices, the inflatable bladder should have a length of approximately 75–100% and a width of 37–50% of the upper-arm circumference. For automated devices, select the cuff size according to the manufacturer’s instructions.

People with large arms, particularly a mid-arm circumference above 42 cm, may need a conical cuff because a rectangular cuff can overestimate pressure. If upper-arm measurement is not possible, a validated electronic wrist-cuff device may be used.

First, the diagnosis should not normally be based on a single office visit unless there is grade 3 hypertension, approximately 180/110 mmHg, or the patient is at high or very high risk because of HMOD or cardiovascular disease. For most patients, accurate office assessment requires at least two or three visits at intervals of one to four weeks, depending on blood-pressure level and cardiovascular risk. The average of the last two of three readings at each visit should be used.

Second, measure pressure one and three minutes after standing to detect orthostatic hypotension in older adults above 65 years, people receiving antihypertensive treatment—especially the very old—people with diabetes, those with neurodegenerative disorders, and anyone with symptoms suggestive of postural hypotension.

At the first visit, measure blood pressure in both arms; simultaneous electronic measurement is preferable when available. If the systolic difference exceeds 10 mmHg, confirm it with repeated measurements. If the difference persists, use the arm with the higher pressure for subsequent measurements because it more accurately reflects pressure in the major arteries and may have better prognostic value. A persistent inter-arm difference above 15–20 mmHg may reflect atherosclerosis or stenosis of large intrathoracic or arm arteries and warrants assessment for arterial disease.

Unattended automated office blood-pressure measurement

Unattended OBPM is performed automatically, usually with three or more readings, while the patient is alone in the examination room and healthcare staff are absent. It aims to standardise the office environment by providing a quiet setting, repeated readings and no conversation.

This method was used in SPRINT, although later retrospective examination suggested that the absence of staff was variable. Available data generally show lower values without staff than with standard OBPM, probably because the alerting response to staff presence and the white-coat effect are reduced. Values may therefore be closer to out-of-office readings, although they do not always agree with home or daytime ambulatory averages.

The main limitation is that evidence linking unattended OBPM to treatment-related outcome reduction is largely restricted to SPRINT, unlike standard OBPM, which is supported by many trials. Evidence that unattended OBPM predicts cardiovascular events and mortality in the general population is also limited. In addition, unattended measurement requires equipment, space and staff, which may make it difficult to accommodate large numbers of patients in primary care or outpatient clinics. The variable difference between unattended and standard OBPM also prevents a reliable correction factor for comparing thresholds and treatment targets across studies.

For these reasons, automated OBPM using a standardised protocol—three readings with the patient in the correct position and conditions—appears to be the most practical and reasonable measurement approach for routine clinical practice.

Blood pressure during exercise

Blood pressure rises during dynamic and static exercise. The increase is more pronounced for systolic than for diastolic pressure, and only systolic pressure can usually be measured reliably by non-invasive methods during exercise. The systolic response is related to pre-exercise resting pressure, age, arterial stiffness and abdominal obesity, and is somewhat greater in men than in women.

Some evidence suggests that an exaggerated exercise response predicts the development of hypertension independently of resting pressure. There is no universally accepted definition of a normal exercise response. A European Association of Preventive Cardiology consensus document has suggested that a peak exercise systolic pressure above 220 mmHg in men or 200 mmHg in women merits further clinical evaluation, including ABPM.

Two observations are of interest: the blood-pressure response to submaximal exercise may have greater prognostic value than the peak response, and a fall in pressure during exercise may indicate cardiovascular disease. Exercise testing is not recommended as a routine part of hypertension assessment because methodologies and definitions are not sufficiently standardised. An exercise-related rise should not discourage people with treated or untreated hypertension from regular exercise, particularly aerobic exercise, unless pressure is extremely high, such as grade 3 hypertension. Regular exercise remains an important lifestyle intervention for sustained blood-pressure reduction.

Blood-pressure measurement in hospital

Because mercury sphygmomanometers are no longer used, hospital wards should use automated mercury-free devices, including digital electronic or hybrid devices. Multiparameter monitors that measure blood pressure, oxygen saturation, temperature and pulse are increasingly common. Some professional monitors offer high-speed measurement in less than 30 seconds or a rapid mode that estimates systolic pressure, which can be useful in emergency departments.

Important features of hospital monitors include programmable repeated measurements at variable intervals and sufficient memory to retain previous readings. All automated monitors require regular maintenance and calibration. At least standard and large adult cuffs, as well as paediatric cuffs, should be available; some devices use a wide-range cuff according to the manufacturer’s instructions. Only devices and cuffs validated according to accepted standards should be used.

In atrial fibrillation, at least three auscultatory readings are recommended because pressure varies from beat to beat. Automated oscillometric devices can also be used: they measure systolic pressure reasonably well and usually only slightly overestimate diastolic pressure. Some devices use an atrial-fibrillation algorithm that can detect the arrhythmia automatically.

Automated oscillometric measurement may also be a reasonable alternative to intra-arterial measurement in intensive-care, resuscitation and perioperative settings. Agreement is acceptable across the normal pressure range in critically ill patients, but underestimation at very low pressures and overestimation at very high pressures have been reported. Underestimation during hypotension is especially concerning because recognising and quantifying hypotension is essential for detecting and preventing vital-organ hypoperfusion. In a meta-analysis comparing oscillometric devices with brachial intra-arterial measurement, systolic pressure was underestimated by an average of 5.7 mmHg at low pressure, with wide variation between patients and devices.

In patients with hypotensive shock, oscillometric mean pressure has been reported to be about 13 mmHg higher than invasive measurements. Although oscillometric devices are often used in emergency and perioperative care, direct arterial catheter measurement should be used in critical illness—especially shock—to guide and titrate vasopressors and fluid therapy.

Central blood pressure

Central, or aortic, pressure can be estimated non-invasively from peripheral pressure-waveforms using tonometry, cuff-based devices and dedicated algorithms. Interest in central pressure reflects three considerations: it is the pressure experienced by vital organs and vessels affected by atherosclerosis; peripheral and central values differ; and treatment effects may differ between them.

A meta-analysis of clinical studies found that central pressure was associated with left-ventricular hypertrophy, carotid intima-media thickening and albuminuria independently of peripheral pressure. Other studies and meta-analyses, however, have produced conflicting results regarding its cardiovascular prognostic value. Higher central pulse pressure has been associated with more cardiovascular events even after adjustment for peripheral systolic pressure.

So far, central measurement has not improved prediction of cardiovascular events or mortality beyond brachial measurement. One study reached a different conclusion and linked high central pressure with cardiovascular and cerebrovascular risk independently of brachial pressure. The incremental prognostic value of central measurement is therefore uncertain. Non-invasive central measurement also has practical limitations and must be calibrated against conventional brachial values, usually obtained by oscillometry.

Despite some reference data, there are no definitive cut-offs distinguishing normal from elevated central pressure and no established agreement between central and brachial pressure across different populations. Widespread use of central measurement in hypertension management cannot therefore be recommended. A possible principal application is isolated systolic hypertension in the young (ISHY), in which peripheral pressure may be disproportionately high despite a normal central pressure. Central assessment may help distinguish a benign “spurious” pattern from ISHY caused by early arterial stiffening. Other waveform-derived measures, such as augmentation index and wave-reflection indices, require further definition before routine clinical use.

Home blood-pressure monitoring

HBPM provides multiple out-of-office readings in a person’s usual environment. It is acceptable for long-term patient use and relatively inexpensive, with costs often borne by the user. Compared with office values, HBPM readings are more reproducible, predict HMOD, cardiovascular events and mortality more effectively, add modest prognostic information when combined with office pressure, measure daily pressure variability, and—like ABPM—identify phenotypes such as masked hypertension (MH) and white-coat hypertension (WCH).

HBPM should use a validated automated upper-arm cuff according to a recognised protocol, such as the listings at STRIDE BP. Devices that store and average multiple readings or transmit data to a phone, computer or secure online system may facilitate clinical review. Conditions and posture should resemble those used for OBPM. Because HBPM usually avoids the white-coat effect, values are lower than office values in most people; the difference becomes smaller as office pressure decreases.

The conventional home threshold corresponding to an office pressure of 140/90 mmHg is approximately 135/85 mmHg, although this is not based on randomised outcome trials. By analogy, the home treatment target should correspond to the recommended office target, which is not precisely defined and is probably only a few mmHg lower. These consensus-based thresholds should be interpreted cautiously because office, home and 24-hour ABPM averages correlate only modestly in treated and untreated people and may differ substantially within the same person.

Record home readings before planned clinic visits or when a clinically important change is suspected. Ideally, monitor for seven days, and not fewer than three days, taking duplicate readings one minute apart in the morning—before medication when treatment is prescribed—and in the evening. Discard the first day, which is often higher and less stable, and average the remaining values.

HBPM can improve the continuity of blood-pressure control during long-term treatment. Limitations include the need for patient training, use of inaccurate devices, anxiety, excessive repeated measurements and patient-initiated treatment changes. Another limitation is the absence of night-time information. This is not trivial because night-time pressure predicts outcomes more strongly than daytime pressure. Newer systems can obtain automated readings during sleep and may provide values, HMOD associations and prognostic information similar to ABPM. As with ABPM, however, outcome trials based specifically on HBPM-guided treatment remain limited.

Recommendations for office or home measurement

• Use a validated automated upper-arm electronic cuff device.
• Measure in a quiet room at a comfortable temperature.
• Avoid smoking, caffeine, food, stimulant or recreational drugs, and exercise for 30 minutes before measurement.
• Sit quietly and relax for 3–5 minutes.
• Do not speak during or between readings.
• Support the back against the chair.
• Keep the legs uncrossed and the feet flat on the floor.
• Rest the bare arm on a table with the middle of the upper arm at heart level.
• In the clinic, take three readings one minute apart and use the average of the last two.
• At home, take two readings one minute apart and use their average.

Ambulatory blood-pressure monitoring

Although the patient must remain still during each measurement, ABPM provides repeated readings under conditions that reflect ordinary life, including daily activity and sleep. It shares the advantages of HBPM over OBPM: greater reproducibility of the 24-hour average, closer association with HMOD and better prediction of outcomes and mortality, as well as identification of WCH and MH.

ABPM can distinguish apparent from true resistant hypertension and quantify features such as 24-hour variability and morning hypertension, which may have adverse prognostic value independent of the 24-hour average. Its most distinctive advantage is quantitative assessment of nocturnal dipping. This is clinically important because night-time pressure predicts events better than daytime pressure; the absence of dipping or a rise in night-time pressure is associated with substantially higher risk. ABPM can also identify episodes of hypotension in daily life and help assess sustained control.

For research on antihypertensive treatment, ABPM has the advantage of little or no placebo effect. It is not suitable for frequent routine use, however: it is relatively expensive, not widely available in primary care, and can disturb sleep. As with HBPM, the major limitation is the lack of randomised trials directly comparing treatment guided by ABPM with treatment guided by OBPM, and the absence of outcome-based thresholds and targets established directly from such trials.

Because ABPM does not produce a white-coat effect, its values are lower than office values. A conventional ambulatory threshold is a 24-hour average of 130/80 mmHg, corresponding approximately to office values of 140/90 mmHg. In the absence of direct outcome trials, treatment targets should be related to the office targets. ABPM uses a fully automated device programmed to record at set intervals over a usual 24-hour day. Patients should record activities, symptoms, meals, medication times, sleep and unusual events. The software provides daytime, night-time and 24-hour averages as well as an hourly profile.

Many recommendations require at least 20 valid daytime readings and seven valid sleep readings over 24 hours. Because the device software may discard readings, hours—especially night-time hours—can be missing. There is also evidence that measurements every 60 minutes, only 24 readings per day, may produce an inaccurate 24-hour average. Measuring every 20 minutes throughout the day and night may reduce the risk that poor data quality cancels the advantages of ABPM and avoids making information scarcest during the most prognostically important period.

The following conventional values define hypertension according to the measurement method: