- Email: [email protected]
Validation of an oscillometric electronic sphygmomanometer in an obstetric population
AJH
1998;11:978 –982
Validation of an Oscillometric Electronic Sphygmomanometer in an Obstetric Population Kenneth Kwek, Yvonne G.S. Chan, Kok Hian Tan, and George S.H. Yeo
Use of an automated electronic sphygmomanometer will allow us to minimize the errors inherent in mercury sphygmomanometry. We conducted this validation according to the 1990 protocol of the British Hypertensive Society. We recruited 87 subjects from the antenatal population of Kandang Kerbau Hospital and took three sequential readings using simultaneously both manual and electronic sphygmomanometry. A total of 261 readings from either method were thus collected and the results analyzed to compare the accuracy of electronically read blood pressure with that assessed manually. We found that 89.9% of the
T
he clinical evaluation of blood pressure is one of the most commonly used screening tests in the antenatal clinic. In most instances, unfortunately there is a large potential for error, as the commonly used Riva-Rocci-Korotkoff method of sphygmomanometry is prone to many potential sources of error. These stem from all aspects of its use, the operator, the instrumentation, and the environment. It becomes obvious then, that if we are to use blood pressure in the diagnosis and management of preeclampsia, first we have to eliminate or at least minimize these sources of error. Proper training in the use and
Received September 4, 1997. Accepted March 9, 1998. From the Department of Maternal Fetal Medicine, Kandang Kerbau Women’s and Children’s Hospital, Singapore, Republic of Singapore. Address correspondence and reprint requests to Dr. Kenneth Kwek, Department of Maternal Fetal Medicine, KK Women’s and Children’s Hospital, 100 Bukit Timah Road, Singapore 229899, Republic of Singapore; e-mail: kenjinpacific.net.sg
© 1998 by the American Journal of Hypertension, Ltd. Published by Elsevier Science, Inc.
electronically read pressures differed from the manually read pressures by 5 mm Hg or less and 98.9% of the electronic readings differed from manual readings by 10 mm Hg or less; only 0.4% of readings had a difference of more than 15 mm Hg. The accuracy of the device was not affected either by the blood pressure or the arm circumference. Am J Hypertens 1998;11:978 –982 © 1998 American Journal of Hypertension, Ltd.
KEY WORDS:
Validation, electronic, sphygmomanometer, obstetrics.
maintenance of the equipment as well as continued monitoring of technique is of the utmost importance to eliminate observer and instrument error. A more attractive alternative is the introduction of automated electronic blood pressure measuring devices. However, it is essential that these instruments are properly calibrated and validated for use in pregnancy. A new highly portable electronic sphygmomanometer (Terumo ES-H51; Terumo Corp., Fuji-City, Sizuoka, Japan), specifically designed for clinical use, has recently become available. The device has previously been validated for clinical use in Japan,1 although only in a heterogeneous group of 64 nonpregnant subjects. Before embracing this new technology in the field of obstetrics, we had to validate the device for use in a pregnant population. METHODS Device The device measures 58 by 21.5 by 92 mm and weighs 104 g (with batteries). A liquid crystal 0895-7061/98/$19.00 PII S0895-7061(98)00093-4
AJH–AUGUST 1998 –VOL. 11, NO. 8, PART 1
ELECTRONIC SPHYGMOMANOMETER IN PREGNANCY
display shows the blood pressure, pulse, pulse wave indicator, usage indicator, and error messages. It runs on a single rechargeable battery, which is good for 80 readings. The blood pressure is measured through a cuff that carries a built-in ceramic microphone, which when placed directly over the brachial artery allows detection of the Korotkoff sounds. The cuff is manually inflated and the device then automatically deflates at a rate of 3 mm Hg/sec. The first Korotkoff sound is taken to mark the systolic blood pressure and the fifth Korotkoff sound is taken for the diastolic pressure. Exclusion of erroneous noise is optimized by synchronization of pulse wave signals and Korotkoff sounds (as extrinsic noises are asynchronous). The device is also able to use the oscillometric method to measure blood pressure, which it does automatically should the auscultatory method fail. The Validation This validation trial was conducted according to the 1990 protocol of the British Hypertensive Society (BHS).2 We chose to use the 1990 protocol for the validation of this electronic sphygmomanometer instead of the more updated 1993 version. This was mainly because in the new protocol, device validation was assessed through the use of sequential same-arm measurements as compared to simultaneous measurements. Simultaneous measurements have been shown to correlate more closely with the measurements of the mercury sphygmomanometer.3 This allows for a more precise validation procedure. Phase I in the BHS protocol calls for strict training and accreditation of nursing staff. In this study, two Registrar-grade observers were recruited for the validation. A pilot trial was conducted on 53 patients in which the two observers simultaneously measured the brachial artery pressure using a two-headed training stethoscope, blinded to each others readings. There was a mean interobserver difference of 2.15 mm Hg, with a range of 0 to 8 mm Hg for systolic readings. For diastolic readings, a mean interobserver difference of 1.74 mm Hg was found with a range of 0 to 6 mm Hg. This showed an acceptable level of interobserver deviation. For phase II, three devices were obtained from the manufacturer and each in turn connected through a Y-connector to a mercury sphygmomanometer. At least 50 pressure readings were simultaneously taken from both devices by two observers through the range of pressure from 20 to 270 mm Hg. In-use assessment of the device in phase III was carried out on three units placed in the Kandang Kerbau Hospital Emergency Room for a 9-month period. During this period, each unit was used on an average of 40 patients per day. Nurses’ comments and patient
979
feedback was sought on the acceptability and user friendliness of the device. For phase IV of the trial, an identical assessment was conducted on the devices previously used in phase III to evaluate interdevice variability after a period of fairly rigorous use. For phase V, one device was arbitrarily selected for the validation trial. This unit was once again connected through a Y-shaped connector to a standard mercury sphygmomanometer. Eighty-seven patients were recruited from the antenatal population of Kandang Kerbau Hospital. The patients were in turn seated and the cuff applied over the left upper arm, ensuring appropriate placement of the microphone over the brachial artery. Arm circumference was measured to ensure use of a suitable cuff. Data on patient age, weight, gestational age, pulse rate, and identity number were also collected. The blood pressure was read simultaneously by the ES-H51 device and a trained observer, using the mercury sphygmomanometer, blinded to the electronic reading. This was performed three times on each patient, giving a total of 261 readings by either method. The pulse and systolic and diastolic pressures, using the fifth Korotkoff sound, were recorded. The difference between manual and electronic readings was then calculated for each measurement pair. The age range of our population was 19 to 41 years with a mean age of 30 years, a slight deviation from the BHS protocol in view of the obstetric population. At least 10% of the patients had at least one blood pressure measurement in the following ranges: diastolic pressure, ,80 mm Hg, 80 to 90 mm Hg, 90 to 100 mm Hg, and .100 mm Hg; systolic pressure, ,130 mm Hg, 130 to 140 mm Hg, 140 to 150 mm Hg, and .150 mm Hg. This too deviated slightly from the BHS protocol in view of the obstetric population under study. Our study population comprised gravid women with a mean gestational age of 32 weeks of amenorrhea, ranging from 20 to 41 weeks of amenorrhea. RESULTS The observers in our trial were two Registrar-grade clinicians with acceptable interobserver variation. Phases II and IV of our trial revealed that the device has a very acceptable interdevice variation both before and after field use. Of the readings taken, .98% were within 2 mm Hg of a simultaneous manual measurement, satisfying the BHS criteria. In-use assessment is more relevant in the evaluation of ambulatory blood pressure monitors. As a result of the feedback we received, the device was found to be user-friendly, hardy, and acceptable to patients. There were few error messages (,1% of all measurements), all of which were remedied by adjusting the cuff to ensure
980
AJH–AUGUST 1998 –VOL. 11, NO. 8, PART 1
KWEK ET AL
TABLE 1. DATA FROM VALIDATION TRIAL SHOWING OBSERVER AND DEVICE MEASUREMENTS
Observer (systolic) Observer (diastolic) Device (systolic) Device (diastolic) Observer–device Systolic difference Observer–device Diastolic difference
Mean
Standard Deviation
Minimum
Maximum
118.7 72.7 115.9 72.8 3.4 2.0
22.6 16.1 23.3 16.6 2.6 1.9
84 43 77 42 0 0
193 119 189 119 17 10
correct placement or use of an appropriately sized cuff. Table 1 shows the distribution of the observer and device measurements as well as the distribution of observer-device difference obtained in phase V of our trial. The trial was conducted according to the BHS protocol for sphygmomanometer validation. Table 2 shows the grading criteria based on the cumulative percentage of readings. As illustrated, the device under validation reached an A grading according to these guidelines, for both systolic and diastolic measurements. Figure 1 shows a scatter plot of the pressure difference (device minus observer) against the mean systolic and diastolic measurements (ie, mean of the device and observer measurements). As can be seen there is no significant correlation between the mean blood pressure measurement and the pressure difference. Similarly, there was no significant relationship between either systolic or diastolic blood pressure difference and arm circumference (Figure 2). It can be concluded from this observer assessment validation trial that the device meets the accuracy criteria as stipulated by the BHS protocol. The device has been shown to be at least as accurate as a trained observer in the determination of blood pressure in a pregnant population.
TABLE 2. GRADING CRITERIA BASED ON THE CUMULATIVE PERCENTAGE OF READINGS (2) AND OBSERVED PRESSURE DIFFERENCES FOR TEST DEVICE Difference Difference Difference <5 mm Hg <10 mm Hg <15 mm Hg Grade A Grade B Grade C Grade D Observer–device systolic difference Observer–device diastolic difference
80% 65% 45% Worse than grade C
90% 85% 75% Worse than grade C
95% 95% 90% Worse than grade C
83.1%
98.1%
99.2%
94.3%
100%
100%
DISCUSSION Preeclampsia is a significant condition in obstetric practice both in terms of its prevalence as well as its potential consequential morbidity and mortality. That the condition has often been labeled a disease of theories is accurate as there is much controversy with regard to almost every aspect of the condition, from the very diagnosis, through the optimal evaluation and determination of the best management strategy. At present, blood pressure is the most commonly applied screening test in the antenatal clinic as current clinical evaluation of preeclampsia relies heavily on the detection and monitoring of raised blood pressure. The use of standardized blood pressure measurement is able to detect a group of patients at high risk of preeclampsia.4 However, this involves a marked increase in personnel and time, as such, in most instances the blood pressure is quantified by the mercury sphygmomanometer using the Riva-RocciKorotkoff sound method. There are, however, many shortcomings in the use of this method, which will in all probability lead to errors in the diagnosis and management of this common condition. These shortcomings stem from all aspects of the method. Observer error stems from the individual involved in the measurement and has several components such as suboptimal visual or auditory acuity, poor concentration or motivation, digit preference, threshold avoidance, and improper interpretation of Korotkoff sounds. Even senior clinicians were shown to have a strong bias with regard to digit preference and the majority failed to use the appropriate size of cuff.5 Instrument error is equally important and may include dirty or broken glass tube, defective stethoscope, inappropriate cuff placement and size, inaccurate calibration of sphygmomanometer, and a noisy environment. Electronic sphygmomanometers allow us to minimise the operator bias and error. The high degree of automation and robustness of these devices make them ideal for clinics with a high patient load requiring a single operator to measure and record a large number of patients’ blood pressures within a certain time constraint. However, it is prudent that such de-
AJH–AUGUST 1998 –VOL. 11, NO. 8, PART 1
ELECTRONIC SPHYGMOMANOMETER IN PREGNANCY
981
FIGURE 1. Scatter plot of systolic and diastolic pressure differences against mean systolic pressure difference.
FIGURE 2. Scatter plot of systolic and diastolic pressure difference against arm circumference.
982
AJH–AUGUST 1998 –VOL. 11, NO. 8, PART 1
KWEK ET AL
vices undergo appropriate validation in the respective population in which they will be used, prior to their introduction into routine clinical practice. In our validation trial, we used the 1990 protocol of the British Hypertensive Society instead of the revised 1993 protocol mainly because it uses simultaneous same-arm measurements. Device and observer blood pressure readings have been found to be closer at simultaneous than at sequential measurements (3), allowing for more accurate validation. We found the device to satisfy the accuracy criteria stipulated by the BHS protocol (2). This is reassuring in that the device is shown to be at least as accurate as a trained observer in the measurement of blood pressure. This device uses the fifth Korotkoff sound for measurement of the diastolic blood pressure. This has been shown to be superior to the use of the fourth Korotkoff sound in that there is less room for subjective interpretation of the measurement (6,7). The oscillometric mode may also be selected, especially in circumstances of high ambient noise, or as the default mode when the auscultatory method fails. Although there were initial reports of inaccuracy possibly leading to increased morbidity (8), oscillometry for blood pressure measurement has already been shown to be an acceptable method of blood pressure determination in pregnancy (9), and several ambulatory sphygmomanometers presently in clinical use use the oscillometric method satisfactorily. There is a tendency to prematurely embrace all new technology in an effort to match the phenomenal rate of technological progress. We must, however, temper this enthusiasm with pragmatism and appropriately evaluate all new technology prior to clinical implementation to ensure that above all, patient care and the cost effective practice of medicine are not compro-
mised. Only after the appropriate clinical trails have proven the superiority and benefit of such technology can we then begin to accept it into our clinical practice. REFERENCES 1.
Imai Y, Hashimoto J, Minami N, et al: Accuracy and performance of the Terumo ES-H51, a new portable blood pressure monitor. Am J Hypertens 1994;7:255–260.
2.
O’Brien E, Petrie J, Littler W, et al: The British Hypertensive Society protocol for the evaluation of automated and semi-automated blood pressure measuring devices with special reference to ambulatory systems. J Hypertens 1990;8:607– 619.
3.
Livi R, Teghini L, Cagnoni S, Scarpelli PT: Simultaneous and sequential same-arm measurements in the validation studies of automated blood pressure measuring devices. Am J Hypertens 1996;9:1228 –1231.
4.
Gallery EDM, Ross M, Hunyor SN, Gyory AZ: Predicting the development of pregnancy-associated hypertension. The place of standardised blood-pressure measurement. Lancet 1977;i:1273–1275.
5.
Perry IJ, Wilkinson LS, Shinton RA, Beevers DG: Conflicting views on the measurement of blood pressure in pregnancy. Br J Obstet Gynaecol 1991;98:241–243.
6.
Shennan AH, Kissane J, de Swiet M: Validation of the SpaceLabs 90207 ambulatory blood pressure monitor for use in pregnancy. Br J Obstet Gynaecol 1993;100:904 – 908.
7.
Halligan A, Shennan A, Thurston H, et al: Ambulatory blood pressure measurement in pregnancy: the current state of the art. Hypertens in Pregnancy 1995;14:1–16.
8.
Quinn M: Automated blood pressure measurement devices: a potential source of morbidity on preeclampsia? Am J Obstet Gynaecol 1994;175:1303–1307.
9.
Shennan A, Halligan A, Gupta M, et al: Oscillometric blood pressure measurement in severe pre-eclampsia: validation of the SpaceLabs 90207. Br J Obstet Gynaecol 1996;103:171–173.
1998;11:978 –982
Validation of an Oscillometric Electronic Sphygmomanometer in an Obstetric Population Kenneth Kwek, Yvonne G.S. Chan, Kok Hian Tan, and George S.H. Yeo
Use of an automated electronic sphygmomanometer will allow us to minimize the errors inherent in mercury sphygmomanometry. We conducted this validation according to the 1990 protocol of the British Hypertensive Society. We recruited 87 subjects from the antenatal population of Kandang Kerbau Hospital and took three sequential readings using simultaneously both manual and electronic sphygmomanometry. A total of 261 readings from either method were thus collected and the results analyzed to compare the accuracy of electronically read blood pressure with that assessed manually. We found that 89.9% of the
T
he clinical evaluation of blood pressure is one of the most commonly used screening tests in the antenatal clinic. In most instances, unfortunately there is a large potential for error, as the commonly used Riva-Rocci-Korotkoff method of sphygmomanometry is prone to many potential sources of error. These stem from all aspects of its use, the operator, the instrumentation, and the environment. It becomes obvious then, that if we are to use blood pressure in the diagnosis and management of preeclampsia, first we have to eliminate or at least minimize these sources of error. Proper training in the use and
Received September 4, 1997. Accepted March 9, 1998. From the Department of Maternal Fetal Medicine, Kandang Kerbau Women’s and Children’s Hospital, Singapore, Republic of Singapore. Address correspondence and reprint requests to Dr. Kenneth Kwek, Department of Maternal Fetal Medicine, KK Women’s and Children’s Hospital, 100 Bukit Timah Road, Singapore 229899, Republic of Singapore; e-mail: kenjinpacific.net.sg
© 1998 by the American Journal of Hypertension, Ltd. Published by Elsevier Science, Inc.
electronically read pressures differed from the manually read pressures by 5 mm Hg or less and 98.9% of the electronic readings differed from manual readings by 10 mm Hg or less; only 0.4% of readings had a difference of more than 15 mm Hg. The accuracy of the device was not affected either by the blood pressure or the arm circumference. Am J Hypertens 1998;11:978 –982 © 1998 American Journal of Hypertension, Ltd.
KEY WORDS:
Validation, electronic, sphygmomanometer, obstetrics.
maintenance of the equipment as well as continued monitoring of technique is of the utmost importance to eliminate observer and instrument error. A more attractive alternative is the introduction of automated electronic blood pressure measuring devices. However, it is essential that these instruments are properly calibrated and validated for use in pregnancy. A new highly portable electronic sphygmomanometer (Terumo ES-H51; Terumo Corp., Fuji-City, Sizuoka, Japan), specifically designed for clinical use, has recently become available. The device has previously been validated for clinical use in Japan,1 although only in a heterogeneous group of 64 nonpregnant subjects. Before embracing this new technology in the field of obstetrics, we had to validate the device for use in a pregnant population. METHODS Device The device measures 58 by 21.5 by 92 mm and weighs 104 g (with batteries). A liquid crystal 0895-7061/98/$19.00 PII S0895-7061(98)00093-4
AJH–AUGUST 1998 –VOL. 11, NO. 8, PART 1
ELECTRONIC SPHYGMOMANOMETER IN PREGNANCY
display shows the blood pressure, pulse, pulse wave indicator, usage indicator, and error messages. It runs on a single rechargeable battery, which is good for 80 readings. The blood pressure is measured through a cuff that carries a built-in ceramic microphone, which when placed directly over the brachial artery allows detection of the Korotkoff sounds. The cuff is manually inflated and the device then automatically deflates at a rate of 3 mm Hg/sec. The first Korotkoff sound is taken to mark the systolic blood pressure and the fifth Korotkoff sound is taken for the diastolic pressure. Exclusion of erroneous noise is optimized by synchronization of pulse wave signals and Korotkoff sounds (as extrinsic noises are asynchronous). The device is also able to use the oscillometric method to measure blood pressure, which it does automatically should the auscultatory method fail. The Validation This validation trial was conducted according to the 1990 protocol of the British Hypertensive Society (BHS).2 We chose to use the 1990 protocol for the validation of this electronic sphygmomanometer instead of the more updated 1993 version. This was mainly because in the new protocol, device validation was assessed through the use of sequential same-arm measurements as compared to simultaneous measurements. Simultaneous measurements have been shown to correlate more closely with the measurements of the mercury sphygmomanometer.3 This allows for a more precise validation procedure. Phase I in the BHS protocol calls for strict training and accreditation of nursing staff. In this study, two Registrar-grade observers were recruited for the validation. A pilot trial was conducted on 53 patients in which the two observers simultaneously measured the brachial artery pressure using a two-headed training stethoscope, blinded to each others readings. There was a mean interobserver difference of 2.15 mm Hg, with a range of 0 to 8 mm Hg for systolic readings. For diastolic readings, a mean interobserver difference of 1.74 mm Hg was found with a range of 0 to 6 mm Hg. This showed an acceptable level of interobserver deviation. For phase II, three devices were obtained from the manufacturer and each in turn connected through a Y-connector to a mercury sphygmomanometer. At least 50 pressure readings were simultaneously taken from both devices by two observers through the range of pressure from 20 to 270 mm Hg. In-use assessment of the device in phase III was carried out on three units placed in the Kandang Kerbau Hospital Emergency Room for a 9-month period. During this period, each unit was used on an average of 40 patients per day. Nurses’ comments and patient
979
feedback was sought on the acceptability and user friendliness of the device. For phase IV of the trial, an identical assessment was conducted on the devices previously used in phase III to evaluate interdevice variability after a period of fairly rigorous use. For phase V, one device was arbitrarily selected for the validation trial. This unit was once again connected through a Y-shaped connector to a standard mercury sphygmomanometer. Eighty-seven patients were recruited from the antenatal population of Kandang Kerbau Hospital. The patients were in turn seated and the cuff applied over the left upper arm, ensuring appropriate placement of the microphone over the brachial artery. Arm circumference was measured to ensure use of a suitable cuff. Data on patient age, weight, gestational age, pulse rate, and identity number were also collected. The blood pressure was read simultaneously by the ES-H51 device and a trained observer, using the mercury sphygmomanometer, blinded to the electronic reading. This was performed three times on each patient, giving a total of 261 readings by either method. The pulse and systolic and diastolic pressures, using the fifth Korotkoff sound, were recorded. The difference between manual and electronic readings was then calculated for each measurement pair. The age range of our population was 19 to 41 years with a mean age of 30 years, a slight deviation from the BHS protocol in view of the obstetric population. At least 10% of the patients had at least one blood pressure measurement in the following ranges: diastolic pressure, ,80 mm Hg, 80 to 90 mm Hg, 90 to 100 mm Hg, and .100 mm Hg; systolic pressure, ,130 mm Hg, 130 to 140 mm Hg, 140 to 150 mm Hg, and .150 mm Hg. This too deviated slightly from the BHS protocol in view of the obstetric population under study. Our study population comprised gravid women with a mean gestational age of 32 weeks of amenorrhea, ranging from 20 to 41 weeks of amenorrhea. RESULTS The observers in our trial were two Registrar-grade clinicians with acceptable interobserver variation. Phases II and IV of our trial revealed that the device has a very acceptable interdevice variation both before and after field use. Of the readings taken, .98% were within 2 mm Hg of a simultaneous manual measurement, satisfying the BHS criteria. In-use assessment is more relevant in the evaluation of ambulatory blood pressure monitors. As a result of the feedback we received, the device was found to be user-friendly, hardy, and acceptable to patients. There were few error messages (,1% of all measurements), all of which were remedied by adjusting the cuff to ensure
980
AJH–AUGUST 1998 –VOL. 11, NO. 8, PART 1
KWEK ET AL
TABLE 1. DATA FROM VALIDATION TRIAL SHOWING OBSERVER AND DEVICE MEASUREMENTS
Observer (systolic) Observer (diastolic) Device (systolic) Device (diastolic) Observer–device Systolic difference Observer–device Diastolic difference
Mean
Standard Deviation
Minimum
Maximum
118.7 72.7 115.9 72.8 3.4 2.0
22.6 16.1 23.3 16.6 2.6 1.9
84 43 77 42 0 0
193 119 189 119 17 10
correct placement or use of an appropriately sized cuff. Table 1 shows the distribution of the observer and device measurements as well as the distribution of observer-device difference obtained in phase V of our trial. The trial was conducted according to the BHS protocol for sphygmomanometer validation. Table 2 shows the grading criteria based on the cumulative percentage of readings. As illustrated, the device under validation reached an A grading according to these guidelines, for both systolic and diastolic measurements. Figure 1 shows a scatter plot of the pressure difference (device minus observer) against the mean systolic and diastolic measurements (ie, mean of the device and observer measurements). As can be seen there is no significant correlation between the mean blood pressure measurement and the pressure difference. Similarly, there was no significant relationship between either systolic or diastolic blood pressure difference and arm circumference (Figure 2). It can be concluded from this observer assessment validation trial that the device meets the accuracy criteria as stipulated by the BHS protocol. The device has been shown to be at least as accurate as a trained observer in the determination of blood pressure in a pregnant population.
TABLE 2. GRADING CRITERIA BASED ON THE CUMULATIVE PERCENTAGE OF READINGS (2) AND OBSERVED PRESSURE DIFFERENCES FOR TEST DEVICE Difference Difference Difference <5 mm Hg <10 mm Hg <15 mm Hg Grade A Grade B Grade C Grade D Observer–device systolic difference Observer–device diastolic difference
80% 65% 45% Worse than grade C
90% 85% 75% Worse than grade C
95% 95% 90% Worse than grade C
83.1%
98.1%
99.2%
94.3%
100%
100%
DISCUSSION Preeclampsia is a significant condition in obstetric practice both in terms of its prevalence as well as its potential consequential morbidity and mortality. That the condition has often been labeled a disease of theories is accurate as there is much controversy with regard to almost every aspect of the condition, from the very diagnosis, through the optimal evaluation and determination of the best management strategy. At present, blood pressure is the most commonly applied screening test in the antenatal clinic as current clinical evaluation of preeclampsia relies heavily on the detection and monitoring of raised blood pressure. The use of standardized blood pressure measurement is able to detect a group of patients at high risk of preeclampsia.4 However, this involves a marked increase in personnel and time, as such, in most instances the blood pressure is quantified by the mercury sphygmomanometer using the Riva-RocciKorotkoff sound method. There are, however, many shortcomings in the use of this method, which will in all probability lead to errors in the diagnosis and management of this common condition. These shortcomings stem from all aspects of the method. Observer error stems from the individual involved in the measurement and has several components such as suboptimal visual or auditory acuity, poor concentration or motivation, digit preference, threshold avoidance, and improper interpretation of Korotkoff sounds. Even senior clinicians were shown to have a strong bias with regard to digit preference and the majority failed to use the appropriate size of cuff.5 Instrument error is equally important and may include dirty or broken glass tube, defective stethoscope, inappropriate cuff placement and size, inaccurate calibration of sphygmomanometer, and a noisy environment. Electronic sphygmomanometers allow us to minimise the operator bias and error. The high degree of automation and robustness of these devices make them ideal for clinics with a high patient load requiring a single operator to measure and record a large number of patients’ blood pressures within a certain time constraint. However, it is prudent that such de-
AJH–AUGUST 1998 –VOL. 11, NO. 8, PART 1
ELECTRONIC SPHYGMOMANOMETER IN PREGNANCY
981
FIGURE 1. Scatter plot of systolic and diastolic pressure differences against mean systolic pressure difference.
FIGURE 2. Scatter plot of systolic and diastolic pressure difference against arm circumference.
982
AJH–AUGUST 1998 –VOL. 11, NO. 8, PART 1
KWEK ET AL
vices undergo appropriate validation in the respective population in which they will be used, prior to their introduction into routine clinical practice. In our validation trial, we used the 1990 protocol of the British Hypertensive Society instead of the revised 1993 protocol mainly because it uses simultaneous same-arm measurements. Device and observer blood pressure readings have been found to be closer at simultaneous than at sequential measurements (3), allowing for more accurate validation. We found the device to satisfy the accuracy criteria stipulated by the BHS protocol (2). This is reassuring in that the device is shown to be at least as accurate as a trained observer in the measurement of blood pressure. This device uses the fifth Korotkoff sound for measurement of the diastolic blood pressure. This has been shown to be superior to the use of the fourth Korotkoff sound in that there is less room for subjective interpretation of the measurement (6,7). The oscillometric mode may also be selected, especially in circumstances of high ambient noise, or as the default mode when the auscultatory method fails. Although there were initial reports of inaccuracy possibly leading to increased morbidity (8), oscillometry for blood pressure measurement has already been shown to be an acceptable method of blood pressure determination in pregnancy (9), and several ambulatory sphygmomanometers presently in clinical use use the oscillometric method satisfactorily. There is a tendency to prematurely embrace all new technology in an effort to match the phenomenal rate of technological progress. We must, however, temper this enthusiasm with pragmatism and appropriately evaluate all new technology prior to clinical implementation to ensure that above all, patient care and the cost effective practice of medicine are not compro-
mised. Only after the appropriate clinical trails have proven the superiority and benefit of such technology can we then begin to accept it into our clinical practice. REFERENCES 1.
Imai Y, Hashimoto J, Minami N, et al: Accuracy and performance of the Terumo ES-H51, a new portable blood pressure monitor. Am J Hypertens 1994;7:255–260.
2.
O’Brien E, Petrie J, Littler W, et al: The British Hypertensive Society protocol for the evaluation of automated and semi-automated blood pressure measuring devices with special reference to ambulatory systems. J Hypertens 1990;8:607– 619.
3.
Livi R, Teghini L, Cagnoni S, Scarpelli PT: Simultaneous and sequential same-arm measurements in the validation studies of automated blood pressure measuring devices. Am J Hypertens 1996;9:1228 –1231.
4.
Gallery EDM, Ross M, Hunyor SN, Gyory AZ: Predicting the development of pregnancy-associated hypertension. The place of standardised blood-pressure measurement. Lancet 1977;i:1273–1275.
5.
Perry IJ, Wilkinson LS, Shinton RA, Beevers DG: Conflicting views on the measurement of blood pressure in pregnancy. Br J Obstet Gynaecol 1991;98:241–243.
6.
Shennan AH, Kissane J, de Swiet M: Validation of the SpaceLabs 90207 ambulatory blood pressure monitor for use in pregnancy. Br J Obstet Gynaecol 1993;100:904 – 908.
7.
Halligan A, Shennan A, Thurston H, et al: Ambulatory blood pressure measurement in pregnancy: the current state of the art. Hypertens in Pregnancy 1995;14:1–16.
8.
Quinn M: Automated blood pressure measurement devices: a potential source of morbidity on preeclampsia? Am J Obstet Gynaecol 1994;175:1303–1307.
9.
Shennan A, Halligan A, Gupta M, et al: Oscillometric blood pressure measurement in severe pre-eclampsia: validation of the SpaceLabs 90207. Br J Obstet Gynaecol 1996;103:171–173.