SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Objective To assess the accuracy of the Omron HEM 705 CP oscillometric device for the measurement of blood pressure in pregnancy and pre-eclampsia.

Setting Two teaching hospitals.

Participants Eighty-five pregnant women with a wide range of blood pressures, and 43 women with pre-eclampsia.

Methods Evaluation was carried out according to the British Hypertension Society protocol, incorporating criteria of the Association for the Advancement of Medical Instrumentation (AAMI).

Results Mean differences (device minus observer) for systolic blood pressure were −0.9 mmHg (SD 10) in normal pregnancy, and −2 mmHg (SD 10) in the women with pre-eclampsia. For diastolic blood pressure, Korotkoff phase V, the differences were −1.5 mmHg (SD 10), and −8 mmHg (SD 8) respectively; i.e. the device consistently under-read the diastolic blood pressure in pre-eclamptic women by a mean of 8 mmHg when compared with conventional sphygmomanometry. Korotkoff phase IV could not be accurately reproduced. According to the British Hypertension Society grading criteria, the device reached a ‘B’ grading in the 85 pregnant women. In pre-eclampsia only a ‘C’ grade was reached for systolic blood pressure, and ‘D’ grade for diastolic blood pressure. Accuracy criteria stipulated by the AAMI were not met in any situation.

Conclusion The Omron HEM 705 CP does not reach acceptable accuracy criteria for blood pressure measurement when compared with trained observers in women with pre-eclampsia, as judged by the British Hypertension Society Protocol. It also failed to meet the AAMI criteria, although the methodology stipulated by the AAMI may not be applicable to a pregnancy population.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

The hypertensive disorders of pregnancy affect more than 10% of the antenatal population1. Each year they complicate 80,000 pregnancies in the United Kingdom and account for between 12% and 24% of antenatal admissions2,3. Hypertensive disorders of pregnancy are one of the common causes of maternal mortality; a high arterial pressure causes intracerebral haemorrhage4. Pre-eclampsia also remains the most common cause of iatrogenic preterm delivery and accounts for 23% of very low birthweight infants (< 1500g)3.

Blood pressure measurement is the most common screening test in pregnancy. There are, however, many problems inherent in using conventional sphygmomanometry5–7. The introduction of automated blood pressure monitors may reduce errors associated with conventional mercury sphygmomanometry, including sampling errors and observer biases such as terminal digit preference and threshold avoidance. Use of these devices also allows ‘white-coat’ hypertension to be excluded, which may improve the relation of blood pressure measurement to outcome. Self initiated blood pressure measurement allows the patient to take a greater role and interest in the management of their hypertension.

Many devices use the oscillometric technique for measuring blood pressure. Oscillatory changes in the artery can be noted from above systole, gradually increasing as the pressure drops, until they reach a maximum at the mean arterial pressure. They then decrease in amplitude until below the diastolic pressure. Functions of the changes in amplitude of the pressure waves are used to describe systolic and diastolic blood pressure. In general manufacturers do not release details of the exact algorithm used to compute these values to avoid industrial piracy.

Before any device can be introduced into clinical practice it requires validation to ensure accuracy. Criteria stipulated by the British Hypertension Society8 and the Association for the Advancement of Medical Instrumentation (AAMI)9 are widely recognised. Due to altered haemodynamics in pregnancy, it is recommended that automated devices are validated specifically in pregnant women8. However further profound changes in haemodynamics can occur in pregnancies complicated by pre-eclampsia10. The accuracy of these machines in pregnancy has been called into question when assessing women with pre-eclampsia11. Oscillometric measurement of blood pressure is dependent on the compliance of the arterial wall. Since pre-eclampsia is associated with increased vasospasm and a reduced circulatory volume, reduced oscillations of the arterial wall may result in the blood pressure being underestimated by automated monitors.

Recently a new home self-initiated (i.e. put on and activated by the patient on each occasion) blood pressure measuring device, the Omron HEM 705 CP has demonstrated good accuracy (British Hypertension Society grade ‘A’ systolic and grade ‘B’ diastolic) in a nonpregnant population12. This monitor uses the oscillometric principle to measure blood pressure. We therefore undertook to assess the accuracy of the Omron HEM 705 CP device for the measurement of blood pressure in pregnant and pre-eclamptic women.

METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

The Omron HEM 705 CP was validated using 85 subjects with a wide range of blood pressures, according to the recommendations of the British Hypertension Society protocol for the evaluation of automated blood pressure monitoring devices. This was extended to determine the AAMI status. The current British Hypertension Society protocol recommends a minimum of 30 pregnant women, but in order to assess accuracy in a meaningful way across the pressure range, we included a larger number. The device was also validated in 43 women with pre-eclampsia as defined according to Redman and Jefferies13: diastolic blood pressure geqslant R: gt-or-equal, slanted 90 mmHg, with an increase from booking blood pressure geqslant R: gt-or-equal, slanted 25 mmHg. This definition only uses blood pressure to define the disease, and was specifically adopted for use in this study to identify women who had undergone pathological changes during pregnancy, and to exclude women with chronic hypertension.

The validation procedure consisted of five phases. Initially two observers were trained and assessed for accuracy against each other and against an expert observer. Phases 2 to 4 involved calibration checks before and after a period of use. Initially, three devices in turn were connected to a calibrated mercury column. The accuracy of each of the three machines was compared on 30 occasions to the mercury column. The British Hypertension Society protocol stipulates that readings from each device should be within 3 mmHg on at least 28/30 occasions when compared with a mercury column.

Each of the three instruments was then used routinely for a one-month period on the day assessment unit, involving at least 400 inflations. The purpose of this phase was to subject the blood pressure device to a period of routine use before performing the main validation procedure. Following this a further calibration procedure was carried out, as above, to ensure that the device had maintained its accuracy.

The main device validation procedure (phase 5) was carried out in 85 pregnant volunteers obtaining informed consent. The study was approved by the local ethics committee. Subjects were selected so that at least 14% of women had their first reading in the following blood pressure ranges: diastolic blood pressure < 80, 80 to < 90, 90 to < 100, geqslant R: gt-or-equal, slanted 100 mmHg; systolic blood pressure < 130, 130 to < 140, geqslant R: gt-or-equal, slanted 140 mmHg. Twelve women had an arm circumference > 35 cm. Only women in the second and third trimester were selected: 30 were in the second trimester and 55 in their third. The two trained observers were assessed after every 20 patients, according to the British Hypertension Society protocol.

Sequential same arm blood pressure measurements were made by the two observers listening simultaneously and then by the device. Seven readings were made in total (four by the observers and three by the device). Readings were taken at one minute intervals. The systolic blood pressure and Korotkoff phases IV (K4) and V (K5) were noted, if present. The two observers and the machine operator were blinded to the measurements of each other by partitions. The best three sequential observer measurements are compared with the device readings to determine its accuracy. The AAMI stipulate that the mean of the differences between the observer and the device be less than ±5 mmHg with a standard deviation leqslant R: less-than-or-eq, slant 8. The British Hypertension Society grade is determined by the percentage of readings within 5, 10 and 15 mmHg. A similar protocol was carried out in the 43 pre-eclamptic women who were recruited from the antenatal wards.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Observer assessment fulfilled the accuracy criteria stipulated by the British Hypertension Society protocol on each of the four occasions performed. The Omron HEM 705 CP passed both the ‘before-use’ and ‘after-use’ device calibration phase by achieving at least 28/30 readings within 3 mmHg of a calibrated mercury column.

Oscillometric devices tend to record systolic blood pressure corresponding to Korotkoff phase I and diastolic blood pressure corresponding to phase V14. We found that phase IV could not be accurately reproduced between our observers: in only 81/340 measurements (24%) did both observers agree to the detection of phase IV, while in 145/340 measurements (42%) both observers agreed that phase IV could not be identified. Therefore validation data for systolic and diastolic blood pressures (phase V) only were collected. Validating against Korotkoff phase IV was thought to be inappropriate as it could not be reproduced even though this phase is recommended as the diastolic identification point in pregnancy by many authorities15–17.

Device validation data showing the mean systolic and diastolic blood pressures (Korotkoff V) taken by observer 1 and observer 2, as well as for device, are shown in Table 1. Results divided into pregnancy (n= 255 paired measurementson 85 pregnant women) and pre-eclampsia (n= 126 paired measurements in 42 pre-eclamptic women). In one subject with pre-eclampsia the Omron HEM 705 CP failed to register a reading after repeated attempts, and this woman was excluded from the analysis (originally 43 women were included). Also shown are the mean and standard deviations of the differences between each observer and the device to determine the result according to the AAMI criteria.

Table 1.  Validation data of the Omron HEM 705 CP in pregnancy and pre-eclampsia for paired measurements (systolic blood pressure and phase V) showing mean blood pressure (K5 = phase V) for observers 1 and 2 and device. Also shown are the mean and standard deviation (SD) of the differences to determine the AAMI status. SBP = systolic blood pressure; DBP = diastolic blood pressure.
  AAMI criteriaReadings (%)*
  MeanDifferenceleqslant R: less-than-or-eq, slant5leqslant R: less-than-or-eq, slant10leqslant R: less-than-or-eq, slant15
  1. *Difference between mmHg readings.

Device validation in 85 pregnant women      
SBPObs 1117(19)0.6(10)527988
 Obs2117(20)0.9(10)557891
 Device116(21)    
DBFObslKS73(14)2(10)487893
 Obs 2 K573(14)1.5(10)507693
 Device72(14)    
Device validation in 42 pre-eclamptic women      
SBPObs 1137(11)4(10)487085
 Obs 2135(11)2(10)487582
 Device132(11)    
DBFObslKS92(10)9(8)335690
 Obs2K592(10)8(8)336182
 Device84(11)    

Table 2 demonstrates how the British Hypertension Society grades are determined by the accumulated percentage agreement within 5, 10 and 15 mmHg. The overall grade is determined by the best observer value. Table 3 shows the grade achieved by the Omron HEM 705 CP in pregnancy and pre-eclampsia based on the Society's grading protocol. The Omron HEM 705 CP achieved an overall British Hypertension Society grade B for systolic blood pressure, and a grade B for Korotkoff phase V in women without pre-eclampsia. The Omron HEM 705 CP failed the AAMI criteria in these women for both systolic blood pressure and diastolic blood pressure (K5). In pre-eclampsia the device achieved an overall grade C for systolic blood pressure and grade D for Korotkoff V. The Omron HEM 705 CP failed the AAMI criteria on all counts.

Table 2.  Grading criteria based on the cumulative percentage of mmHg readings, following the British Hypertension Society protocol for blood pressure measuring devices.
 Difference between observer and test device (%)
Gradeleqslant R: less-than-or-eq, slant5leqslant R: less-than-or-eq, slant10leqslant R: less-than-or-eq, slant15
A608595
B507590
C406585
D Worse than C 
Table 3.  The grade achieved by the Omron HEM 705 CP in pregnancy and pre-eclampsia based on the British Hypertension Society's grading protocol for blood pressure measuring devices. SBP = systolic blood pressure; DBP = diastolic blood pressure; V = Korotkoff phase.
  Difference between observer and test device (%) 
  leqslant R: less-than-or-eq, slant5leqslant R: less-than-or-eq, slant10leqslant R: less-than-or-eq, slant15Grade
In pregnancy     
Obs 1SBP527988C
 DBP(V)487893C
Obs2SBP557891B
 DBP(V)507693B
In pre-eclampsia     
Obs 1SBP487085C
 DBP(V)335682D
Obs 2SBP487590C
 DBP (V)336182D

Figures 1 and 2 show Bland-Altman plots18 of the pressure differences (device minus observer) plotted against systolic and diastolic blood pressures (calculated as the mean of the readings of the device and the observer) for the best observer and for the Omron HEM 705 CP in women without pre-eclampsia (n= 255). Figures 3 and 4 show similar Bland-Altman plots for systolic and diastolic blood pressure (phase V) for the best observer in those women with pre-eclampsia (n= 126). Lines have been drawn to demonstrate the points within 5, 10 and 15 mmHg difference. The percentage of points within these lines determines the British Hypertension Society grade. The Bland-Altman plots demonstrates the change in accuracy (device minus observer value) with the variation in mean blood pressure. In Fig. 4 it can be seen that most of the points lie below the zero x-axis indicating that the machine consistently under-reads Korotkoff phase V in women with pre-eclampsia. Therefore, Korotkoff phase IV is likely to have an even greater error.

image

Figure 1. Bland-Altman plot of pressure difference in mmHg (device minus best observer) against systolic blood pressure in 85 women without pre-eclampsia (n= 255).

Download figure to PowerPoint

image

Figure 2. Bland-Altman plot of pressure difference in mmHg (device minus best observer) against diastolic blood pressure in 85 women without pre-eclampsia (n= 255).

Download figure to PowerPoint

image

Figure 3. Bland-Altman plot of pressure difference in mmHg (device minus best observer) against systolic blood pressure in 42 women with pre-eclampsia (n= 126).

Download figure to PowerPoint

image

Figure 4. Bland-Altman plot of pressure difference in mmHg (device minus best observer) against diastolic blood pressure in 42 women with pre-eclampsia (n= 126).

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Using current guidelines, the Omron HEM 705 CP cannot be recommended to detect or monitor pre-eclampsia since it achieved a C grade for the systolic blood pressure and a D grade for the diastolic blood pressure, and failed the AAMI criteria on all counts. Because of consistent bias, in pre-eclampsia an Omron diastolic blood pressure of 82 mmHg is likely to be equivalent to a blood pressure (Korotkoff V) of 90 mmHg when measured by conventional sphygmomanometry.

The AAMI criteria have been derived for simultaneous measurements, and have not been tailored particularly for pregnancy. Given the sequential nature of the British Hypertension Society analysis, and the inherent variability of blood pressure in pregnancy, it is possible that the AAMI criteria are not applicable to our population19. Further work is needed to ascertain whether variability of blood pressure in pregnancy is similar to that in a nonpregnant population before current criteria can be applied to pregnancy. This is particularly important because this device achieved an A grade for the systolic blood pressure and a B grade for the diastolic blood pressure in the nonpregnant population. Our work demonstrates the necessity of separate validation specifically in pregnancy and pre-eclampsia. It also suggests that new guidelines may need to be established to assess the accuracy of machines in pregnancy. Although this device cannot be recommended by current criteria, the validation procedure itself may not be applicable to a pregnant population.

The use of automated blood pressure monitors in the pregnant population has come under scrutiny because of their widespread use by obstetric personnel. Some investigators have found large discrepancies between oscillometric devices and auscultation by observers, particularly in pre-eclampsia. Both the Dinamap 1840 SX (Critikon, Tampa, Florida, USA)20 and the Accutorr III (Datascope, Mondale, New Jersey, USA)11 significantly under-read the diastolic blood pressure. Both these devices are commonly used on labour wards to monitor high risk patients. However, the methodology used in these studies was not as stringent as the guidelines recommended by the British Hypertension Society and the AAMI, and Korokoff phase IV was used in the Accutorr III study thereby increasing the apparent error. To our knowledge there is only one device, the SpaceLabs 90207 (Redmond, Washington, USA) which has demonstrated reasonable accuracy in pre-eclampsia21. This device in fact differs from the Omron in that it is an ambulatory monitor which inflates automatically by a pre-determined interval and is not removed by the patient. Before any blood pressure monitor can be used in pregnancy it must be validated specifically in pregnancy including patients with severe pre-eclampsia. Currently this is not the case in the British Hypertension Society or AAMI recommendations.

To date, very few studies have been conducted to compare the accuracy of direct intra-arterial measurements with indirect auscultatory methods in hypertensive pregnancies. There is only one study which compared Korotkoff phase IV to intra-arterial measurements in hypertensive pregnancies and demonstrated an average under-recording of 16 mmHg10. Those studies that have investigated normotensive pregnancies demonstrate an under recording of Korotkoff phase V by 4–7 mmHg22,23 and considerably more for Korotkoff phase IV24. The algorithms for oscillometric blood pressure devices are derived from intra-arterial studies on anaesthetised patients. Given that intra-arterial blood pressure has been shown to be lower than that determined by conventional sphygmomanometry in pregnancy, it may be that inaccurate automated devices actually measure closer to the true intra-arterial blood pressure. Conventional sphygmomanometry may be inappropriate for the validation of these devices and therefore new definitions with respect to pregnancy and pre-eclampsia should be sought. Further studies comparing conventional sphygmomanometry, automated blood pressure devices and intra-arterial measurements in pre-eclamptic women must be carried out to clarify this matter. If we are to accurately identify women at risk of cerebral haemorrhage these devices may be better than conventional sphygmomanometry. Considering the advantages of home blood pressure monitoring, outcome data would have to be related to oscillometric determination of blood pressure in the home environment before these measurements could be correctly interpreted, though theoretically they may improve the value of blood pressure measurement.

References

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References
  • 1
    National High Blood Pressure Education Program Working Group Report on High Blood Pressure in Pregnancy. Am J Obstet Gynecol 1990; 163: 16911712.
  • 2
    Twaddle S, Harper V. An economic evaluation of daycare in the management of hypertension in pregnancy. Br J Obstet Gynaecol 1992; 99: 459463.
  • 3
    Rosenberg K, Twaddle S. Screening and surveillance of pregnancy hypertension—an economic approach to the use of daycare. Baillieres Clin Obstet Gynaecol 1990; 4: 89107.
  • 4
    Report on Confidential Enquiries into Maternal Deaths in the United Kingdom 1988–1990. London : HMSO, 1994.
  • 5
    Perry IJ, Wilkinson LS, Shinton RA, Beevers DG. Conflicting views on the measurement of blood pressure in pregnancy. Br J Obstet Gynaecol 1991; 98: 241243.
  • 6
    Perry IJ, Stewart BA, Brockwell J, Khan M, Davies P, Beevers DG et al. Recording diastolic blood pressure in pregnancy. BMJ 1990; 301: 1198.
  • 7
    Brown MA, Simpson JM. Diversity of blood pressure recording during pregnancy: implications for the hypertensive disorders. Med J Aust 1992; 156: 306308.
  • 8
    O'Brien E, Petrie J, Littler W, de Swiet M, Padfield PL, Altman DG, et al. An outline of the revised British Hypertension Society protocol for the evaluation of blood pressure measuring devices. J Hypertens 1993; 11: 677679.
  • 9
    American National Standard. Electronic or automated sphygmomanometers. Arlington , Virginia : AAMI: 1992.
  • 10
    Wallenburg HCS. Hemodynamics in hypertensive pregnancy. In: RubenPC, editor. Hypertension in pregnancy. Amsterdam : Elsevier, 1988: 66101.
  • 11
    Quinn M. Automated blood pressure measurement devices: a potential source of morbidity in preeclampsia Am J Obstet Gynecol 1994; 170: 13031307.
  • 12
    O'Brien E, Mee F, Atkins N. An accurate automated device for home blood pressure measurement at last! The Omron HEM 705 CP. Proceedings of the British Hypertension Society; Dublin ; 1994 [abstract].
  • 13
    Redman CW, Jefferies M. Revised definition of pre-eclampsia. Lancet 1988; 1: 809812.
  • 14
    Korotkoff NS. A contribution to the problem of the methods for the determination of blood pressure. Reports Imperial Military Medical Academy 1905; 11: 365.
  • 15
    Petrie JC, O'Brien ET, Littler WA, de Swiet M. Recommendations on blood pressure measurement. BMJ 1986; 293: 611615.
  • 16
    Davey DA, MacGillivray I. The classification and definition of the hypertensive disorders of pregnancy. Am J Obstet Gynecol 1988; 158: 892898.
  • 17
    World Health Organisation Study Group. The hypertensive disorders of pregnancy. WHO Technical Report. 1987; Geneva : World Health Organisation. Series No. 758:.
  • 18
    Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 307310.
  • 19
    Atkins N, Mee F, O'Malley K, O'Brien E. The relative accuracy of simultaneous same arm, simultaneous opposite arm and sequential same arm measurements in the validation of automated blood pressure measuring devices. J Hum Hypertens 1990; 4: 647649.
  • 20
    Marx GF, Schwalbe SS, Cho E, Whitty JE. Automated blood pressure measurements in laboring women: are they reliable Am J Obstet Gynecol 1993; 168: 796798.
  • 21
    Shernnan A, de Swiet M, Halligan A. Accuracy of oscillometric blood pressure measurements in preeclampsia. Am J Obstet Gynecol 1995; 172: 13251326.
  • 22
    Brown MA, Reiter L, Smith B, Buddle ML, Morris R, Whitworth JA. Measuring blood pressure in pregnant women: a comparison of direct and indirect methods. Am J Obstet Gynecol 1994; 171: 661667.
  • 23
    Raftery EB, Ward AP. The indirect method of recording blood pressure. Cardiovasc Res 1968; 2: 210218.
  • 24
    Johenning AR, Barron WM. Indirect blood pressure measurement in pregnancy: Korotkoff phase 4 versus phase 5. Am J Obstet Gynecol 1992; 167: 577580.