Blood pressure measuring devices: recommendations of the European Society of HypertensionBMJ 2001; 322 doi: https://doi.org/10.1136/bmj.322.7285.531 (Published 03 March 2001) Cite this as: BMJ 2001;322:531
- Eoin O'Brien (), professor of cardiovascular medicinea,
- Bernard Waeber, professorb,
- Gianfranco Parati, associate professorc,
- Jan Staessen, lecturerd,
- Martin G Myers, professore the European Society of Hypertension Working Group on Blood Pressure Monitoring.
- a Blood Pressure Unit, Beaumont Hospital, Dublin 9, Ireland
- b Centre Hospitalier Universitaire Vaudois, Division D'Hypertension, Departement de medecine interne, 1011 Lausanne, Switzerland
- c Universita degli Studi di Milano-Biocca, Istituto Scientifico Ospedale San Luca, IRCCS, Istituto Auxologico Italiano, Via Spagnoletto 3, 20149 Milan, Italy
- d Katholieke Universiteit Leuven, Hypertensie en Cardiovasculaire Revalidatie Eenheid, Inwendige Geneeskunde-Cardiologie, UZ Gasthuisberg, 3000 Leuven, Belgium
- e Division of Cardiology, Sunnybrook and Women's College Health Sciences Centre, 2075 Bayview Avenue, Toronto, ON, Canada M4N 3M5
- Correspondence to: E O'Brien
- Accepted 17 November 2000
There is a large market for blood pressure measuring devices not only in clinical medicine but also among the public where the demand for self measurement of blood pressure is growing rapidly. For consumers, whether medical or lay, accuracy should be of prime importance when selecting a device to measure blood pressure. However, most devices have not been evaluated for accuracy independently using the two most widely used protocols: the British Hypertension Society (BHS) protocol and the standard set by the US Association for the Advancement of Medical Instrumentation (AAMI). 1 2 The Working Group on Blood Pressure Monitoring of the European Society of Hypertension has decided to review blood pressure measuring devices regularly to guide purchasers.3 For this first report devices for which there is published evidence of independent validation using these protocols have been surveyed. Because most blood pressure devices have not been independently validated, only a fraction of the many devices available have been surveyed. Devices that have been validated recently for which results have not yet been published were not included, but this shortcoming should be addressed in future.
Two manual sphygmomanometers have been validated, one is recommended
Five devices for clinical use in hospitals have been validated, two are recommended
23 devices for self measurement of blood pressure have been validated, five are recommended
24 devices for ambulatory measurement of blood pressure have been validated, 16 are recommended
Validations and recommendations will be updated on the BMJ's website
In 1987, the American Association for the Advancement of Medical Instrumentation published a standard for sphygmomanometers which included a protocol for evaluating the accuracy of devices.4 In 1990 a protocol was devised by the British Hypertension Society.5 Both protocols have since been revised. 1 2 Since the two protocols can be reconciled the joint criteria are applied in most validation studies.6 The criteria for fulfilling the BHS protocol are that devices must achieve at least grade B (where A denotes greatest agreement with mercury standard and D denotes least agreement) for systolic and for diastolic pressures (table 1)1; the criteria for fulfilling the AAMI protocol are that the test device must not differ from the mercury standard by a mean difference >5 mm Hg or a standard deviation >8 mm Hg.2
Criteria for recommendation
The following criteria have been used to designate devices according to accuracy. A device is classed as recommended if it fulfils the AAMI criteria for both systolic and diastolic pressures (denoted as passed) and received a grade of A or B under the BHS protocol for both systolic and diastolic blood pressures. A device is not recommended if it fails the AAMI criteria for either systolic or diastolic pressure and achieves a grade of C or D for either systolic or diastolic pressure under the BHS protocol. A questionable recommendation is made when there is doubt about the strength of evidence. This may occur when a device fulfils the criteria of one protocol but not the other, and it may be best not to recommend the device for clinical use until a confirmatory study has been performed; when the validation results were presented only in an abstract without sufficient detail to appraise the methodology, and it may be best to withhold an opinion until the full results have been published or at least provided to a potential purchaser by the manufacturer; when the conditions of the protocols have not been fully adhered to (listed as a protocol violation in our scheme); or when a device fulfils the AAMI criteria for intra-arterial validation (the BHS protocol does not advocate validation using direct intra-arterial measurement1), but it may be best to await a validation against indirect blood pressure measurement before recommending the device for general clinical use.
Identification of devices
This review was based on two previous surveys (which should be consulted for early validation studies that are not reproduced in this review), 7 8 and computerised search programs were used to identify validation studies in the literature published up to December 1999. Blood pressure measuring devices were divided into two broad categories: manual sphygmomanometers, which include mercury and anaeroid devices; and automated sphygmomanometers, which include devices for clinical use in hospitals, for self measurement of blood pressure, for ambulatory blood pressure measurement, and for measuring blood pressure in community settings. (Information on manufacturers appears on the BMJ's website.)
These devices are listed in table 2. 9-12 One model of the many mercury sphygmomanometers available, the PyMah (PyMah Corporation, Flemington, NJ), has been validated according to both protocols and is recommended.9 Because mercury sphygmomanometers are generally of a simple basic design with standard components, it is probably reasonable to assume that most would be of similar accuracy. The standard anaeroid sphygmomanometer has only been formally validated according to the calibration procedure of the BHS protocol,12 and the results support reservations about anaeroid devices because they are susceptible to becoming inaccurate without this being apparent to the user.13
Devices for use in hospitals
Devices for self measurement of blood pressure
There are a large number of automated devices for self measurement of blood pressure, virtually all of which use the oscillometric technique. Formerly these devices used automated inflation and deflation of a cuff applied to the upper arm over the brachial artery. Recently this technique has been used to measure blood pressure over the radial artery at the wrist, but since these devices become inaccurate if the arm is not kept at heart level during measurement, the working group is reluctant to recommend them regardless of accuracy.19 Devices that measure blood pressure by occluding a digital artery in the finger are also available, but because the problem of limb position is even more critical and there is the additional problem of peripheral vasoconstriction affecting accuracy, this technique is no longer recommended. These devices have not been considered in this review.19
Automated devices for upper arm measurement are shown in table 4. 9 20-29 Automated devices for wrist measurement are listed in table 5.29-32 These have been validated against brachial arterial measurements.
Devices for ambulatory measurement of blood pressure
There are two techniques for measuring ambulatory blood pressure: the commonly used method of intermittent measurement over 24 hours and the developing method of continuous waveform analysis.
Intermittent measurement—Devices dependent on intermittent measurement are listed in table 6.33-68 Many of these devices have been validated for use in specific groups, such as elderly people and pregnant women, and in differing circumstances, such as during exercise and in different postures. Validating devices for use in ambulatory conditions presents many methodological difficulties, and some evidence suggests that inaccuracies found during static conditions may be amplified in ambulatory conditions.69
Devices for continuous non-invasive blood pressure monitoring of a finger—The Portapres (TNO, Amsterdam), a portable recorder for 24 hour ambulatory monitoring, can provide beat to beat monitoring which gives waveform measurements similar to intra-arterial recordings. However, the technique is subject to various inaccuracies, which the use of correction factors and digital filters in the latest model may remove; this model is awaiting formal validation. 70 71
Devices for measuring blood pressure in a community setting
There is little information available on the accuracy of automated devices that are installed in public retail areas—such as pharmacies, supermarkets, health clinics, and companies in a variety of industries—which permit the public to measure blood pressure without charge in an unsupervised, crowded setting with high ambient noise. Evaluations of one such device, the Vita Stat 90550 (Spacelabs Medical, Redmond, WA), have had conflicting results. 72 73
Manufacturers of blood pressure measuring devices use innovative technology to provide an array of systems that can analyse, store, and display features of a haemodynamic variable in ways that would have been beyond the dreams of the pioneers of the technique. Although the selection of a blood pressure measuring device may be influenced by many factors, a fundamental requirement must be that it gives accurate measurements; too often accuracy has been sacrificed for technological ingenuity.
The evidence from validation studies is accumulating, and devices are being scrutinised more critically; this has been the case with ambulatory devices used in specific populations, such as in children, elderly people, and pregnant women. However, the evidence is not always clear cut. There may have been protocol violations; the data published may have been inadequate, such as sometimes occurs when only abstracts have been published; and there may be disagreement between validation studies of the same device. None the less certain recommendations can be made to assist potential purchasers.
In interpreting the recommendations made by this survey the following factors should be considered. A device fulfilling the AAMI criteria and graded A or B for both systolic and diastolic pressure under the BHS protocol has been recommended on grounds of accuracy without equivocation; one that fails the AAMI protocol for either systolic or diastolic pressure and has a grade of C or D for either systolic or diastolic pressure under the BHS protocol cannot be recommended on the grounds of accuracy. Devices are given a questionable recommendation if there is an element of doubt in interpreting the results of a validation study. One circumstance that a purchaser should also consider, but for which we cannot make a recommendation, is the occasional conflict that arises when a device fulfils the criteria of the protocols when validated at one centre but not another. When this occurs the details of the methodology may need to be scrutinised to determine if differences in the selection of participants, for example, might explain the conflict; it may be best to await the results of a confirmatory study before deciding whether the device is accurate.
Only a fraction of the devices available worldwide have been independently validated. This is especially true of devices used for self measurement. In 1994, Ng and Small surveyed 423 automated devices, of which 161 were designed for self measurement.74 Since then the number of devices available for self measurement has increased greatly but comparatively few have been validated. The situation is even worse for automated devices designed for use in specialised areas of hospitals, such as operating theatres and intensive care units, where accuracy should be a priority. Only five of the hundreds of devices available have been validated using the two protocols, of which only two met the criteria for recommendation in this review. 8 75 The much used anaeroid sphygmomanometer has only recently been independently evaluated.12 However, because these devices become inaccurate with use without this inaccuracy being apparent to the user, it is also necessary to validate them after they have been used for some time.13
A serious dilemma is how to influence manufacturers to modify devices that have been shown to be inaccurate. The Dinamap Portable Monitor model 8100 (Critikon, Tampa, FL) is an example of this: despite a number of reports of inaccuracy 18 76 it is one of the most popular automated devices used in clinical practice and hypertension research. It seems that purchasers and users of expensive devices for blood pressure measurement in specialised hospital areas are prepared to accept the word of manufacturers with regard to their accuracy and performance and to ignore warnings from the scientific literature as to their shortcomings.
Again, the situation is worse for self measurement devices. Despite the poor accuracy record of devices that measure blood pressure at the wrist and the serious misgivings voiced by clinicians about these devices,19 their popularity is growing. In Germany, for example, 1.2 million self measurement devices are sold annually.30
The European Union and international organisations of specialists in hypertension have unanimously recommended that all devices for measuring blood pressure should be independently validated. 19 77-82 The reality is, however, that most devices are not validated independently. This may be partly due to the expense of conducting validation studies using complex protocols. 1 2 Recently, the European Society of Hypertension Working Party on Blood Pressure Monitoring agreed proposals to simplify the BHS protocol without compromising the integrity of the procedures, and an international protocol for validation is being drafted.83 This will help manufacturers to market devices worldwide, expedite validation procedures, reduce the expense of performing studies, and permit more centres to undertake validation procedures; all of which would enable manufacturers to have all devices validated independently before they are marketed. Early publication of validation studies might further encourage manufacturers to have their devices evaluated, and the readiness of Blood Pressure Monitoring to act as a repository of peer reviewed studies is welcomed. The internet might provide a means of continuously updating information on blood pressure measuring devices.
Competing interests EO'B is director of the blood pressure unit at Beaumont Hospital, which has been contracted by manufacturers from time to time to conduct validation studies of blood pressure measurement devices; the results of these studies have been published. EO'B has advised AccuSphyg, which is developing a non-mercury automated device for use in hospitals, and he holds a minority financial interest in the company.
Additional information about manufacturers and members of the working group appear on the BMJ's website
Members of the working group are Roland Asmar, Andrew Coats, Denis L Clement, Peter de Leeuw, Robert Fagard, Yutaka Imai, Jean-Michel Mallion, Giuseppe Mancia, Thomas Mengden, Martin Myers, Eoin O'Brien, Paul Padfield, Gianfranco Parati, Paolo Palatini, Thomas G Pickering, Josep Redon, Jan Staessen, Peter Trenkwalder, Gert van Montfrans, Paolo Verdecchia, Bernard Waeber, William White