Intended for healthcare professionals

Rapid response to:


In praise of mercury sphygmomanometers

BMJ 2000; 321 doi: (Published 16 December 2000) Cite this as: BMJ 2000;321:1534

Rapid Response:

Selecting the appropriate sphygmomanometer

EG Lawes rightly draws attention to the problems associated with the
gradual disappearance of the mercury sphygmomanometer,1 referring to
Professor O'Brien's earlier editorial.2 This topic needs further active

Since the editorial appeared, the Medical Devices Agency published a
recommendation that users should consider alternatives to mercury
devices.3 However, there still appears to be some confusion over the
advantages and disadvantages of the alternatives. The mercury
sphygmomanometer when used by trained staff is still classed as the 'gold
standard', but it is now fairly well accepted that the health and safety
and environmental issues associated with mercury will lead to its eventual
disappearance. Aneroid devices are also in widespread use, but because
they are mechanical devices they can be knocked out of calibration easily.
There is no reason why they should not continue to be used, provided the
manufacturers' recommendation to calibrate them every six months is
followed, but indications are that this advice is rarely taken.

Automated devices are now readily available. Initially, they often
had unacceptably poor accuracy, and Professor O'Brien and his colleagues
are to be congratulated for drawing attention to this. The first standard
for accuracy was contained in the publication by the Association for the
Advancement of Medical Instrumentation in 1992.4 It set a minimum
standard, which many devices failed to reach. This document is technical,
and its requirements are not always clear to clinical users. For a device
to be acceptable, you should expect that when the errors in comparison
with a standard mercury device are averaged over all your patients, the
mean errors for systolic and diastolic pressures should be no worse than 5
mmHg, and that the standard deviation of these errors should be no worse
than 8 mmHg. Using this information the British Hypertension Society
introduced a protocol which made these numbers easier to understand in
daily practice - for a device to be acceptable no more than 25% of
measurements should be in error by more than 10 mmHg and no more than 10%
by 15 mmHg.5

Automated devices have a well accepted role in monitoring for changes
in blood pressure, but a clinical decision needs to be made about their
acceptability for other uses. The combined recommendation of the European
Society of Cardiology, European Society of Hypertension and European
Atherosclerosis Society is quite clear - automated devices are "unsuitable
as a routine substitute for the measurement of clinic blood pressure in
the diagnosis of
hypertension" and are "not appropriate for determining the need for
treatment and for assessing treatment efficacy".6

We should not allow the argument that clinical staff are poor at
taking manual measurements to influence us. Clinical staff can be
trained.7 The evaluation protocol of the British Hypertension Society
requires clinical observers to agree within 5 mmHg for 90% of
measurements,5 and measurements should be read to the nearest 2 mmHg.8

I work with others on the European Standards Committee for
sphygmomanometers, and we have recently heard voices of discontent about
the current low level of clinical accuracy required by the Standard for
automated devices. Some would like to see significant improvements, but
that will not happen in the near future. Manufacturers will resist this
strongly, simply because better accuracy cannot yet be achieved and, as
Professor O'Brien points out in his editorial, the "oscillometric
techniques cannot measure blood pressure in all situations".2 The solution
is for users to decide when automated devices are appropriate and
when they are not.

The looming difficulties over blood pressure measurement have been
clear for some years. Recognising this, at Freeman Hospital and Newcastle
University in Newcastle upon Tyne, we developed a manual device in
collaboration with a manufacturer of traditional sphygmomanometers. This
device uses modern electronic methods to produce an accurate alternative
to the mercury sphygmomanometer, with features to improve the measurement
technique of the user. This now introduces another manual sphygmomanometer

I would welcome feedback on the current blood pressure measurement
device standards. Standards can help by weeding out devices which do not
perform as required by the standard, since they should not receive the CE
mark, but beware of devices supplied as an aid to exercise, as they may
not be classified or evaluated as medical devices.3 Remember that
standards do not make recommendations on which devices should be used. In
practice, a clinical decision must be made to select the device
appropriate to the clinical need. This discussion is important and must

1 Lawes EG. In praise of mercury sphygmomanometers. BMJ 2000;

2 O'Brien E. Replacing the mercury sphygmomanometer. BMJ 2000; 320:815-

3 MDA. Blood pressure measurement devices - mercury and non-mercury
(MDA DB2000(03)). London: Medical Devices Agency, 2000.

4 AAMI. Electronic or automated sphygmomanometers (ANSI/AAMI SP10-
1992). Arlington VA: Association for the Advancement of Medical
Instrumentation, 1992.

5 O'Brien E, Petrie J, Littler W, de Swiet M, Padfield PL, Altman DG,
et al. The British Hypertension Society protocol for the evaluation of
blood pressure measuring devices. J Hypertension 1993; 11(suppl 2): S43-

6 Wood D, De Backer G, Faergeman O, Graham I, Mancia G, Neil A, et al.
Clinician's manual on total risk management. London: Science Press, 2000.

7 Beevers M, Beevers DG. Blood pressure measurement in the next
century; a plea for stability. Blood Pressure Monitoring 1996; 1(suppl

8 British Hypertension Society. Blood pressure measurement CD ROM.
London: BMJ Books, 1999.

Competing interests: No competing interests

03 January 2001
Alan Murray
Professor of Cardiovascular Physics
Freeman Hospital, Newcastle upon Tyne