It is well established that numerous errors, biases and omissions in recording blood pressure exist. This study had two objectives. Firstly, to measure the accuracy of sphygmomanometers used in general practice and secondly to assess digit bias in blood pressure recording. This study was carried out in the then Northfield PCG, which comprised 18 practices and 67 GPs. A total of 131 mercury and aneroid sphygmomanometers were tested for accuracy by a trained technician in accordance with the methods specified in BS 2743 (1990). Accuracy was defined as an error of greater than 10 mm Hg. The second part of the methodology involved undertaking an audit of the proportion of registered patients aged 35–80 years who had their blood pressure measured within the last 5 years by members of the Primary Health Care Team. The results were that of the mercury and aneroid sphygmomanometers tested, 17% were inaccurate. Of these, 4% recorded an error greater than 10 mm Hg. One percent of mercury and 10% of aneroid sphygmomanometers recorded an error greater than 10 mm Hg respectively. Sixteen (12%) sphygmomanometers were so deteriorated (air leaks, dirt in mercury) that the researcher suggested their immediate withdrawal from service. The results of the blood pressure recording audit suggested digit bias of both systolic and diastolic recordings to the nearest 10 mm Hg. This study suggests that sphygmomanometers used in general practice are very likely to be inaccurate and some may well be so deteriorated that they should be withdrawn from service. The results of the blood pressure audit showed digit bias in systolic and diastolic readings to the nearest 10 mm Hg. The implications for clinical care—both over diagnosis and under diagnosis—although not assessed are likely to be appreciable. PCG Clinical Governance teams in conjunction with Practice Clinical Leads must address these basic issues.
Blood pressure is an important CHD risk factor and predictor of life expectancy. Guidelines published by the British Hypertension Society recommend that all adults should have their blood pressure measured routinely at least every 5 years until the age of 80 years. Those with high blood pressure should be re-measured annually.1 Treatment of high blood pressure reduces strokes, heart attack and heart failure.2 In order to obtain accurate blood pressure measurement, it is therefore important to ensure that an adequately calibrated and accurate device is used and that staff are trained. The accuracy of the blood pressure reading is of great clinical importance as errors may deprive hypertensive patients of the benefits of hypertension treatment or expose normotensive patients to unnecessary treatment.3,4,5,6 Furthermore, having taken patients blood pressure accurately it is important to ensure that it is recorded correctly.
In this article we estimate the accuracy of mercury and aneroid sphygmomanometers used in general practice. We also report the findings of an audit of people aged 35–80 years old who had their blood pressure measured and recorded within the last 5 years in order to investigate whether significant recording bias in blood pressure readings exist.
Patients and methods
The study was undertaken in Northfield PCG, which was then a level one PCG in South Birmingham, UK. It was set in a suburban surrounding with a population of approximately 110 000. It had 18 practices with two branch surgeries, one single-handed practice and 67 GPs. The demographic make up, levels of service provision and other performance indicators suggest that it was typical of many other PCGs in England at the time. As such we feel that the findings presented here will be similar to those found in many other PCG's in England.
Fifteen out of the 18 practices accepted an invitation to have the accuracy of mercury and aneroid sphygmomanometers used by practice staff checked by a person trained in calibrating sphygmomanometers. The accuracy of mercury and aneroid sphygmomanometers was checked in accordance with the methods specified in BS 2743 (1990). The British Hypertension Society recommends that the accuracy of blood pressure readings should be ‘within 2 mm of mercury’.1
All 18 practices took part in the blood pressure audit. A representative group sample size based on a statistically approved method (95% confidence intervals ± 2.5% error) was calculated for each practice.7 One of the authors (AS) trained a member of each practice to select records in accordance with the general principle of random selection. This was by selecting every ‘xth’ record, where ‘x’ was found by dividing the study group size (ie all 35–80 year olds on list) by the sample size. A member of the practice staff systematically obtained medical records and recorded the last systolic and diastolic blood pressure readings together with the age of the patient onto a data collection sheet. AS entered this data into the database for analysis.
A total of 139 sphygmomanometers were identified and tested. Of these, 92 (66%) were mercury, 39 (28%) aneroid and eight (6%) electronic. Accuracy of only mercury and aneroid sphygmomanometers was tested and the results are shown in Table 1. Electronic sphygmomanometers were not tested. This was because discussions with Omron, the manufacturer's of nearly all electronic sphygmomanometers, revealed that it would not be practical to calibrate electronic sphygmomanometers as part of this study.
An appreciable number of both mercury and aneroid sphygmomanometers showed errors, which the British Hypertension Society would find unacceptable. One percent mercury and 10% aneroid sphygmomanometers recorded an error of greater than 10 mm Hg respectively. Four percent of both types of sphygmomanometers recorded an error greater than 10 mm Hg. Sixteen (12%) sphygmomanometers were so deteriorated (air leaks, dirt in mercury, etc) that the tester suggested their immediate withdrawal from service.
Figures 1 and 2 show blood pressure reading digit bias for systolic and diastolic blood pressures respectively. These suggest that readings are recorded by staff to the nearest 10 mm Hg rather than the recommended 2 mm Hg by the British Hypertension Society. For instance we would expect each of the last digits to be recorded on 10% of occasions. In fact it was found that 0’s are usually vastly over-represented occurring on 75% of records.
Results of this study are consistent with other published research. Bailey et al8 found 35% of aneroid manometers to be inadequately calibrated. Burke et al9 found 30% of aneroid devices with a magnitude of error >4 mm Hg compared with 2% of mercury devices. Mion et al10 found 58% aneroid and 21% of mercury manometers to be poorly calibrated. Our results suggest that 17% of mercury and aneroid sphygmomanometers are inaccurate. Four percent showed an error greater than 10 mm Hg. These inaccuracies could clearly lead to errors in blood pressure measurement and have major consequences for the appropriate treatment of the patient.
The second part of this study found significant digit bias in systolic and diastolic blood pressure readings. The British Hypertension Society recommends that the accuracy of blood pressure readings should be ‘within 2 mm of mercury’. The results of our study show that there is little evidence that this advice is being followed. It appears that blood pressures are likely to be rounded to the nearest 10 mm mercury. Also we noted a slight tendency to round to the nearest 5 mm Hg and even numbers.
These findings suggest that many GPs are rendering less than optimal care to patients attending for general medical services. We are unsure of the legal implications for the PCG (or GP), but suspect these could be profound, because it could be maintained that diagnostic inaccuracy and damage could result from (or at least is compounded) the fact that a GP has not made any arrangements to have his sphygmomanometers calibrated. This is despite the fact that nearly all GPs in the PCG were aware that sphygmomanometers become inaccurate with use. Therefore, we recommend that the Boards of PCG/Ts put into place a mechanism to ensure that a regular check of the accuracy of sphygmomanometers is undertaken. The setting up of a maintenance service contract with a medical supplier can ensure annual calibration and accuracy in blood pressure readings11. Quite clearly the cost of such a contract is one which will have to be borne by GPs.
We recommend that practices implement a quality assurance service based on the 7R's which are as follows:
Recalibrate all sphygmomanometers used in the practice;
Remove unreliable instruments;
Replace them with reliable instruments and provide a written
Record which shows that this is being done Regularly and Reliably by a Responsible person.
The use of semi-automated blood pressure monitors approved by the British Hypertension Society is a possible solution to the problem. However, it is important that the semi-automated devices are regularly calibrated and checked for accuracy until such a time that we are confident that such devices are reliable.
Arrangements must be made by GPs to ensure reliability of their sphygmomanometers. It may be appropriate for PCG/Ts to coordinate and facilitate this. As it is, we believe that very many sphygmomanometers used in general practice in England are inaccurate. We recommend that an annual maintenance service contract with a medical supplier is negotiated and that staff are trained on recording blood pressure accurately.
Ramsey LE et al. Guidelines for management of hypertension: report of the third working party of the British Hypertension Society J Hum Hypertens 1999; 13: 569–592
Markandu ND, Whitcher F, Arnold A, Carney C . The mercury sphygmomanometer should be abandoned before it is prescribed J Hum Hypertens 2000; 14: 31–36
National Institute of Health, National Heart, Lung, and Blood Institute. The Sixth Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure Arch Intern Med 1997; 153: 154–183
Zanchetti A et al. The 1993 Guidelines for the management of mild hypertension: memorandum from WHO/ISH meeting Hypertension 1993; 22: 392–403
Perloff D et al. Human blood pressure determination by sphygmomanometry Circulation 1993; 88: 2460–2470
Frohlich E et al. Recommendations for human blood pressure determination by sphygmomanometers Circulation 1988; 77: 501A–514A
Derry J . Sample size for audit Managing Audit in General Practice, Summer 1993: 17–20
Bailey RH, Knaus VL, Bauer JH . Aneroid sphygmomanometers Arch Intern Med 1991; 15: 1409–1412
Burke MJ et al. Sphygmomanometers in hospital and family practice: problems and recommendations Br Med J 1982; 285: 469–471
Mion D Jr, Pierin AMG . How accurate are sphygmomanometers? J Hum Hypertens 1998; 12: 245–248
We are grateful to all the GPs and staff in Northfield PCG for their co-operation and support in this study. We would also like to thank Mr David Stenson, Chief Officer, and Ann Rouine, Clinical Governance Nurse Board member, for their helpful comments on the manuscript.
About this article
Cite this article
Ali, S., Rouse, A. Practice audits: reliability of sphygmomanometers and blood pressure recording bias. J Hum Hypertens 16, 359–361 (2002). https://doi.org/10.1038/sj.jhh.1001384
- blood pressure measurement
- blood pressure recording bias
JACC: Heart Failure (2020)
BMC Medical Research Methodology (2020)
Results of a project to calibrate mercury sphygmomanometer blood pressure-measuring devices in Egypt
Journal of Human Hypertension (2020)
Journal of Glaucoma (2019)
Home Blood Pressure Telemonitoring: Rationale for Use, Required Elements, and Barriers to Implementation in Canada
Canadian Journal of Cardiology (2017)