AJH 2000;13:276 282 Measurement Variation Among 12 Electronic Home Blood Pressure Monitors Steven A. Yarows and Robert D. Brook Although there are AAMI and BHS standards for accuracy of electronic home electronic blood pressure monitors (HBPM), patient composition differences and differences in manufacturer s algorithm for calculation of the systolic and diastolic measurement may result in measurement differences between monitors. The aim of this study was the measurement of differences among HBPM. Paired comparisons were performed between the Omron 712c electronic home monitor and each of 12 other HBPM (Sunbeam 7654, Sunbeam 7623, Omron 711, Omron 432c, A&D-UA767, Lumiscope 1085M, Omron 725CIC, Assure A30, Lumiscope 1083N, Omron 815, Omron 605, and Assure BD- W20), in addition to comparison to the auscultatory method by trained observers. Measurements were made in normotensive subjects in an ambulatory setting. The main outcome measures were systolic and diastolic blood pressure measurements. All of the HBPM, except for the Sunbeam 7654 and the Assure A30/ BD-W20 (wrist) models, demonstrated small differences of <4/4 mm Hg for systolic/diastolic measurements with pulse measurement differences of <3 beats/min. These differences were less than the differences previously reported for office BP auscultation of 6/5 10 mm Hg for systolic/diastolic measurements. The Omron 712c, passing previous AAMI and BHS standards, measured the systolic reading within 2 mm Hg of auscultatory mercury or aneroid measurement and under-measured the diastolic by 6 9 mm Hg. Differences in the patient composition studied could account for the difference. The wrist and finger manometers performed clinically similar to the Omron 712c, except for the Assure BD-W20, which overmeasured the diastolic by 7 mm Hg. It is concluded that the small differences among the various HBPM, which are less than those in clinical office auscultation, should encourage greater use of electronic manometers in the office and at home. Am J Hypertens 2000;13:276 282 2000 American Journal of Hypertension, Ltd. KEY WORDS: Blood pressure, self-care, home blood pressure devices, validation, standards. Hypertensive patients commonly use home blood pressure monitors, and electronic oscillometric monitors are increasingly being used in hospitals and outpatient settings to measure blood pressure (BP), although there is usually little physician input into the purchase. 1 3 Digital electronic monitors are also being used in hypertensive outcome-based studies. 4 Home blood pressure monitor (HBPM) types include aneroid and electronic models. Most modern electronic HBPM Received April 12, 1999. Accepted July 19, 1999. From the Division of Hypertension, Internal Medicine, University of Michigan Health System, Ann Arbor, Michigan. An educational grant from Bristol-Myers-Squibb was used for the purchase of the electronic devices. Address correspondence to Steven A. Yarows, MD, FACP, Chelsea Internal Med. Consultants PC, 128 Van Buren Street, Chelsea, MI 48118; e-mail: syarows@umich.edu 2000 by the American Journal of Hypertension, Ltd. 0895-7061/00/$20.00 Published by Elsevier Science, Inc. PII S0895-7061(99)00182-X
AJH MARCH 2000 VOL. 13, NO. 3, PART 1 MEASUREMENT VARIATION AMONG HOME BP MONITORS 277 use the oscillometric method to determine BP at the upper arm, wrist, or finger. Some have disparaged the use of wrist and finger monitors as having uncertain accuracy and marked postural variability. 5 7 Each of the manufacturers of HBPM uses a unique proprietary algorithm to calculate systolic and diastolic BP. 8 There are many distributors of electronic blood pressure devices, which results in many different model names, although many distributors use similar manufacturer models. The Association for the Advancement of Medical Instrumentation (AAMI) and the British Hypertensive Society Working Party on Blood Pressure Measurement (BHS) standards have been established to assure accuracy of the HBPM. 9 11 The AAMI and BHS standards both require the monitors to be tested on 85 different subjects with defined characteristics. Manufacturers must comply with the AAMI standards to distribute and sell HBPM in the United States, however, the Food and Drug Administration (FDA) does not actually review the data from the AAMI test, nor are these results usually published in peer-reviewed journals. Thus, physicians have only to assume that the proper tests have been performed. Because each monitor that is studied using AAMI/ BHS standards has its own subject composition, different monitors could use different types of subjects, which could potentially result in measurement differences among different HBPM. We wanted to determine the magnitude of these differences, as this has not been previously studied. METHODS The HBPM were purchased in retail stores for the study, with the exception of the Omron 725CIC (Omron Healthcare, Inc., Vernon Hills, IL) and the Omron 605 (wrist), which were donated by pharmaceutical companies (Novartis and Bristol-Myer Squibb, respectively) for patient use. One author (R.B.) previously owned the aneroid sphygmomanometer. All BP measurements were performed on two authors (S.Y. and R.B.). Two different Omron 712C monitors were used in the study. A paired comparison of the Omron 712C upper arm manometer to another HBPM, mercury sphygmomanometer, or aneroid sphygmomanometer was performed. Each paired comparison was performed only by one subject and was independently performed of the other comparisons. Each monitoring was performed in the seated position with the back supported, always using the left arm. A total of 85 readings were performed on each HBPM monitor and, as each monitor was always compared to either the Omron 712c or a mercury/aneroid manometer, each paired set of readings had 170 BP measurements. Ten readings were first obtained at a single setting with the Omron 712C, and the next 10 readings were obtained with the comparison monitor. The beginning test monitor sequence was alternated to prevent any measurement bias due to the order of readings. The measurements were not separated by any specific interval and often were obtained immediately after the last reading. The final measurement set of each monitor contained five readings to complete the 85 measurements. The aneroid or mercury sphygmomanometer readings were performed by each of the authors on the other author. The aneroid readings were also performed by a licensed practical nurse (LPN) that has practiced clinical internal medicine for 10 years. The wrist and finger manometers were tested with the wrist placed on a surface, which resulted in the top of the wrist horizontal to 5 inches below the superior border of the left clavicle (heart level). The measurements were entered into Microsoft Excel 97 (Microsoft Corporation, Redmond, WA). Standard deviation, 95% confidence intervals, two-sample equal variance two-tailed Student t test, mean average, and differences among the simulation settings and the machines were calculated using these functions in Excel. All of the Omron and A&D devices were calibrated for static measurements (80, 110, 120, and 150 mm Hg) of accuracy using the BP Pump (Bio-Tek Instruments, Winooski, VT). The machine was last calibrated May 6, 1998, and was to be due for recalibration in May 1999. Nether the wrist nor finger HBPM could be calibrated by any known method. The other HBPM were calibrated with the BP Pump using dynamic measurements of 120/80 and 110/70 mm Hg through the previously described technique. 12 The aneroid manometer was tested with the BP pump at 150, 120, 110, and 80 mm Hg and the difference was 1 mmhg. RESULTS Ninety-Five Percent Confidence Intervals of the Paired Measurements The 12 devices tested were all sold for home BP measurement. The devices were compared to the Omron 712c, the calibrated aneroid manometer, and/or to a mercury manometer. A lack of any statistically significant difference was seen in the comparisons of the Omron 712c and other Omron devices for both systolic and diastolic measurement, except for the systolic comparison of the Omron 725CIC and diastolic comparison of the Omron 432c, Omron 725CIC, and Omron 815 (finger). The Sunbeam 7654, Lumiscope 1085M, Assure A30, Lumiscope 1083N, and Assure W20 demonstrated significant differences in both systolic and diastolic measurements; however, the Sunbeam 7623 and A&DUA767 did not have significant differences in either measurement. The pulse measurements were concordant for all devices except the Omron 712c versus the Lumiscope 1083N (finger) and the Omron 605 (wrist).
278 YAROWS AND BROOK AJH MARCH 2000 VOL. 13, NO. 3, PART 1 FIGURE 1. The systolic, diastolic, and pulse differences of the electronic device minus Omron712c (BPM). All are corrected for a single Omron712c device by adding or subtracting the measured average difference of the other Omron712c device. Correction for the Two Omron 712c Machines There were two different Omron 712c machines used for the different paired measurements. Author R.B. had 1530 BP measurements performed and author S.Y. had 2040 measurements performed for a total of 3570 BP measurements in this study. Each Omron 712c machine was compared another other using the same technique of 85 readings for each machine. The first Omron 712c machine read the systolic higher by 1.4 mm Hg and the diastolic higher by 1.5 mm Hg than the second Omron 712c machine (P.05 for both). The second Omron 712c measured the pulse 0.4 beats/ min higher than the first Omron 712c (P.4). Although the differences between the two Omron 712c machines were not clinically significant, the measurements were adjusted to correct for the differences between machines. If the second Omron 712c machine was used, adding 1.4 and 1.5 mm Hg to the respective systolic and diastolic readings corrected the differences in the systolic and diastolic measurements to compare to the first Omron 712c measurements. Similarly, 0.4 beats/min was subtracted from the second Omron 712c readings to compare to the first Omron 712c machine. Differences Between HBPM The adjusted differences between the devices are seen in Figure 1. All of the HBPM had a small diffrence between the HBPM and Omron 712c of 4 mm Hg for the systolic measurements, excluding Sunbeam 7654 (6 mm Hg difference). Self-reading the aneroid systolic measurements demonstrated a larger difference of 7 to 12 mm Hg compared to the automatic Omron 712c. There were small differences ( 1.3 mm Hg) between the Omron 712c and the aneroid or mercury manometers if another trained observer took the systolic measurements. All of the HBPM demonstrated small differences when compared to the Omron 712c in diastolic measurement of 4 mm Hg with the exception of the Assure BD-W10 wrist (over-measured by 7 mm Hg), and Assure BD A30 (over-measured by 7 mm Hg). When the Omron 712c device was compared to the aneroid or mercury manometer, the Omron 712c was
AJH MARCH 2000 VOL. 13, NO. 3, PART 1 MEASUREMENT VARIATION AMONG HOME BP MONITORS 279 TABLE 1. PREVIOUSLY PUBLISHED METHODS FOR COMPARISON OF DIFFERENT HBPM Author, Ref. No. of Subjects BP/Subject No. of Models No. of BP/Models Van Egmond et al 17 30 100 30 100 O Brien-Pre-Home 18 10 6 6 models (3 monitors each) 6 (2 monitors for 3 models) O Brien-Device Validation 18 85 3 4* 225 Bortolotto 30/model 3 2 90 Imai-Simultaneous 20 83 120/model 2 3 8 243 313 4 models (3 monitors each) 90 Imai-Inter-device 20 13 83 Evans 19 28 12 23 36 Yarows/Brook 2 2040 and 1530 12 85 * Four of the initial seven models passed the pre-home and after-use variability tests. BP, blood pressure; HBPM, home blood pressure monitors. found consistently to under-measure the diastolic pressures. Self-measurement of the diastolic pressure using an aneroid manometer showed the Omron 712c to under-measure the diastolic by 2 to 11 mm Hg. Measurement of the diastolic by the other author using a mercury manometer demonstrated an undermeasurement of 6 to 9 mm Hg with a similar 9 mm Hg under-measurement by the LPN using an aneroid manometer. All devices showed a clinically insignificant difference of 3 beats/min of the pulse measurement compared to the Omron 712c. DISCUSSION What is the best technique to compare measurement differences between HBPM? Most articles that determine accuracy of HBPM test only a single device against a standard device. 13 16 There are limited comparisons using multiple HBPM (Table 1). 17 21 Different methods were used for the comparisons in these studies with a variety of subject numbers (13 to 120 per model), number of models compared (2 to 30), BP per subjects (3 to 100), and number of BP per models (6 to 100). As there is no established gold standard, we chose to use two normotensive subjects, so as to limit human variation, using 85 readings per model to minimize the measurement errors. This study was not meant to validate the accuracy of the HBPM tested but, rather, to compare the monitors. The development of AAMI and BHS standards has improved the accuracy of the monitors. The HBPM that were manufactured before AAMI or BHS standards usually failed these tests. 18,22 Most manufacturers except for Omron Healthcare, Inc, do not publish the monitor testing in peer-reviewed journals. The following monitors have passed the AAMI and the BHS standards: Omron 711, Omron 722c, Omron 713c, Omron 403c, Terumo ES-H51, Assure A30, Assure W20, A&D UA-767, and A&D UA-767PC. 23 29 The Omron 706 failed the BHS standards by only receiving a B grade by BHS standards for systolic and a C grade for diastolic measurements. Lumiscope and Sunbeam did not make AAMI or BHS reports available to us. Our main finding was that there were only small differences by all the HBPM (Figure 1), which would unlikely be clinically important. The second important finding is a constant under-measurement of the diastolic measurement by all of the HBPM when compared to measurements by trained personnel using the auscultatory technique. The third clinically relevant finding was that, contrary to previous reports, the wrist and finger manometers tested produced similarly precise BP readings comparable to upper arm HBPM when used correctly. The Omron 712c usually measured the systolic slightly lower than the other devices, whereas the diastolic was both under- and over-measured compared to the other monitors. Generally, there were smaller differences between different Omron models and the Omron 712c than other manufacturers. The paired pulse measurements were similar, with the greatest difference ( 3 beats/min) measured between the Omron 712c and the Lumiscope 1083N (finger). Thus, generally the sensors are equivalent at sensing the pulse, however, the different algorithms used by the different manufacturers probably account for the small differences in the measurement of systolic and diastolic readings. The largest difference between the Omron 712c and the electronic device was over-measurement of the systolic by 6 mm Hg (Sunbeam 7654), diastolic by 7 mm Hg (Assure-BD-W20 wrist), and pulse by 3 beats/ min (Lumiscope 1083N finger). This measurement error is similar to the human error measured in an actual clinical, non research setting. The measurement error in a Brazil University Hospital was 6 mm Hg for
280 YAROWS AND BROOK AJH MARCH 2000 VOL. 13, NO. 3, PART 1 systolic and 10 mm Hg for diastolic, even after taking a test on proper BP measurement technique. 30 A trained registered nurse was compared to usual healthcare providers at a family practice center, with the majority of the healthcare providers being medical assistants and licensed practical nurses, along with a few registered nurses. 31 A difference of 6 mm Hg systolic and 5 mm Hg diastolic was measured. Potentially correctable errors that could be improved by training were estimated to be only 2 mm Hg for systolic and 1 mm Hg for diastolic. Although the Assure A30 measured the greatest difference compared to the Omron 712c by over-measuring the systolic by 4 mm Hg and diastolic by 7 mm Hg, this difference was comparable to the human error in measurement. The human error is variable with each BP measurement. If the same electronic device is used, most of the difference would be due to BP variation rather than machine error, as the difference in measurement that we observed would be constant because of the algorithm defining systolic and diastolic BP and the low variability of the machines. 12 The Omron 712c very accurately measured the systolic pressure within 2 mm Hg of the measurement by trained observers using an aneroid or mercury manometer. Manually inflating a cuff was previously reported to increase the systolic BP by 12 to 13 mm Hg because of the required muscular activity. 32 Self-reading the systolic measurements by the aneroid method demonstrated a7to12mmhgdifference compared to the automatic Omron 712c; however the two manually inflated electronic machines (Sunbeam 7623 and Omron 432c) were easy to inflate and only demonstrated a 2 mm Hg difference compared to the fully automated Omron 712c. The Omron 712c under-measured the diastolic reading by 6 to 9 mm Hg when compared to an aneroid or mercury manometer measured by a trained observer. The Omron 711 and the Omron 725CIC, when compared to a mercury manometer, under-measured the diastolic by 7 mm Hg (P.01) and 6 mm Hg (P.05), respectively. 33 The Omron R3 wrist manometer was also found to under-measure diastolic compared to mercury measurement; however, there was no significant difference when compared to intra-arterial measurements. 34 This would suggest that the diastolic measurement algorithm with Omron monitors better reflects diastolic intra-arterial than auscultatory pressure. The Omron HEM705CP, which uses the same protocol as the Omron 712c, achieved a B systolic grade and an A diastolic grade using the British Hypertensive Society (BHS) protocol. 35,36 The Omron HEM705CP passed the AAMI accuracy criteria with the systolic under-measured by 2 7 mm Hg and the diastolic under-measured by 1 6 mm Hg. The difference of this study with our results is unclear; however, this study used two normotensive subjects, and the AAMI test was performed in mostly normotensive subjects, along with additional hypertensive and hypotensive patients with a variety of arm circumferences. Additionally, the AAMI population was European versus American in this study. Measuring BP at the wrist or finger requires the measurement point to be placed at the heart level to avoid systemic error from hydrostatic pressure. 5 Raising the wrist to a vertically up position lowered the BP by 13/18 mm Hg compared to heart level, whereas lowering the wrist to a vertically down position artificially raised the BP by 18/10 mm Hg. We carefully measured the wrist and finger BP at heart level and found that the readings were clinically similar to the systolic measurement of the Omron 712c. The wrist and finger measurements were also similar to the diastolic measurement of the Omron 712c, except that the Assure BD-W20 (wrist) over-measured the diastolic by 7 mm Hg. The Omron HEM-601 (wrist) was compared by hourly measurements to the boso Oscillomat (upper arm) and SpaceLabs 90207 ambulatory BP machines and the results were similar to our study. 37 No significant differences were recorded between the daily mean systolic (P.07) or diastolic (P.97) BP determined by ambulatory measurements. The Ormon HEM-601 also passed an AAMI test. 38 The following finger devices did not pass AAMI or BHS standards: Omron HEM-812F, Marshal F-88, EW276H (National Ltd), and Omron HEM804F. 39 42 Increased systolic pressure in a geriatric population increased the difference between the finger and upper arm measurement technique, although this may not be clinically important. 41 A limitation of this study is that BP was only measured in two subjects that were both normotensive. Comparison of different electronic monitors using hypertensive and hypotensive patients should be performed. Nonetheless, accuracy of measurement is most important between 130 and 150 mm Hg systolic and between 80 and 100 mm Hg diastolic, as clinical decisions are determined based on these measurements at these levels. It is reassuring that each machine had small differences between each other. The small differences between the machines were equal to or less than the difference that is seen in the usual clinical measurement by auscultation. The small differences between wrist/finger and upper arm machines may encourage increased patient usage, if patients can be given instruction about the correct method of usage. If the same electronic manometer brand is used, the instrument error will be less than measurement with auscultation. Although the diastolic measurement for the
AJH MARCH 2000 VOL. 13, NO. 3, PART 1 MEASUREMENT VARIATION AMONG HOME BP MONITORS 281 HBPM was lower than the mercury measurement in our study, the recommended upper limit of normal for home BP measurement is also lower (135/85 mm Hg) and this should correct for this difference. 43 Additionally, home BP readings over a 2-week period were similar to office readings when the same model of HBPM was used for both measurements. 4 Electronic manometers should be encouraged for wider clinical usage both in the office and at home. ACKNOWLEDGMENTS We are grateful to Judy Williams, LPN, for her assistance in the measurement of the aneroid blood pressures. We are also grateful for the careful review by Alan Weder, MD. REFERENCES 1. Jeck T, Edmonds D, Mengden T, Schubert M, Vetter W: [Performing self-measurement of blood pressure: a patient survey] Umgang mit der Blutdruck-selbstmessung: eine Patientenumfrage: Schweiz Rundsch Med Prax 1991;80:456 461. 2. Personal communication, Eric Vennemeyer, Frost & Sullivan, 1997. 3. Krecke HJ, Fleischmann C, Bokmann M: [Distribution and acceptance of self measurement of blood pressure in the Hamburg area] Verbreitung und Akzeptanz der Blutdruckselbstmessung im Grossraum Hamburg. Schweiz Rundsch Med Prax 1989;78:1336 1342. 4. Hansson L, Zanchetti A, Carruthers SG, Dahlof B, Elmfeldt D, Julius S, Menard J, Rahn KH, Wedel H, Westerling S: Effects of intensive blood-pressure lowering and low-dose aspirin in patients with hypertension: principal results of the Hypertension Optimal Treatment (HOT). Lancet 1998;351:1755 1762. 5. Pickering TG: Technologic advances in home blood pressure monitoring. Blood Press Monit 1998;3(suppl 1):S3 S6. 6. Nesselroad JM, Flacco VA, Phillips DM, Kruse J: Accuracy of automated finger blood pressure devices. Fam Med Mar 1996;28:189 192. 7. Latman NS, Latman A: Evaluation of instruments for noninvasive blood pressure monitoring of the wrist. Biomed Instrum Technol 1997;31:63 68. 8. Yong P, Geddes LA: A surrogate arm for evaluating the accuracy of instruments for indirect measurement of blood pressure. Biomed Instrum Technol 1990;24:130 135. 9. Electronic or automated sphygmomanometers. Developed by Association for the Advancement of Medical Instrumentation. Approved October 16, 1992. American National Standards Institute, Inc.; 330 Washington Boulevard, Suite 400 Arlington, Virginia 22201-4598:1 40. 10. O Brien E, Petrie J, Littler W, et al: The British Hypertension 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. 11. O Brien E, Petrie J, Littler W, et al: Short report: an outline of the revised British Hypertension Society protocol for the evaluation of blood pressure measuring devices. J Hypertens 1993;11:677 679. 12. Yarows SA, Amerena JV: Accuracy of 10 models of home blood pressure monitoring using an oscillometric simulator. What is 120/80? Blood Pressure Monitor :45 52. 13. James GD, Yee LS, Cates EM, Schlussel YR, Pecker MS, Pickering TG: A validation study of the Instromedix Baro-Graf Qd home blood pressure monitor. Am J Hypertens 1990;3:717 720. 14. Foster C, McKinlay S, Cruickshank JM, Coats AJS: Accuracy of the Omron HEM 706 portable monitor for home measurement of blood pressure. J Hum Hypertens 1994;8:661 664. 15. Mufunda J, Sparks B, Chifamba J, Dakwa C, Matenga JA, Adams JM, Sparks HV: Comparison of the Omron HEM-713C automated blood pressure monitor with a standard ausculatory method using a mercury manometer. Cent Afr J Med 1996;42:230 232. 16. Walma EP, van Dooren C, van der Does E, Prins A, Mulder P, Hoes AW: Accuracy of an oscillometric automatic blood pressure device: the Omron HEM403C. J Hum Hypertens 1995;9:169 174. 17. Van Egmond J, Lenders JWM, Weernink E, Thien T: Accuracy and reproducibility of 30 devices for selfmeasurement of arterial blood pressure. Am J Hypertension 1993;6:873 879. 18. O Brien E, Mee F, Atkins N, O Malley K: Inaccuracy of seven popular sphygmomanometers for home measurement of blood pressure. J Hypertens 1990;8:621 634. 19. Evans CE, Haynes RB, Goldsmith CH, Hewson SA: Home blood pressure-measuring devices: a comparative study of accuracy. J Hypertens 1989;7:133 142. 20. Imai Y, Abe K, Sasaki S, Minami N, Munakata M, Sakuma H, Hashimoto J, Sekino H, Imai K, Yoshinaga K: Clinical evaluation of semiautomatic and automatic devices for home blood pressure measurement: comparison between cuff-oscillometric and microphone methods. J Hypertens 1989;7:983 990. 21. Borotolotto LA, Henry O, Hanon O, Sikias P, Mourad JJ, Girerd X: Validation of two devices for self-measurement of blood pressure by elderly patients according to the revised British Hypertension Society protocol: the Omron HEM-722c and HEM-735c. Blood Press Monit 1999;4:21 25. 22. Harrison DW, Kelly PL. Home health-care: accuracy, calibration, exhaust, and failure rate comparisons of digital blood pressure monitors. Med Instrum 1987;21: 323 328. 23. Personal communication, Accuracy of the Omron HEM-711 portable monitor for home measurement of blood pressure. Omron Healthcare Inc, Osamu Shirasaki, September 2, 1998. 24. Cordoba R, Fuertes MI, Alvarez A, Molina I, Solans R, Melero I: [The evaluation of a self-measurement arterial pressure monitor: the OMRON-HM 722C]. [Spanish] Atencion Primaria 1997;20:247 250. 25. Mufunda J, Sparks B, Chifamba J, Dakwa C, Matenga JA, Adams JM, Sparks HV: Comparison of the Omron HEM-713C automated blood pressure monitor with a
282 YAROWS AND BROOK AJH MARCH 2000 VOL. 13, NO. 3, PART 1 standard ausculatory method using a mercury manometer. Cent Afr J Med 1996;42:230 232. 26. Walma EP, van Dooren C, van der Does E, Prins A, Mulder P, Hoes AW: Accuracy of an oscillometric automatic blood pressure device: the Omron HEM403C. J Hum Hypertens 1995;9:169 174. 27. Kwek K, Chan Y, Tan KH, Yeo G: Validation of an oscillometric electronic sphygmomanometer in an obstetric population. Am J Hypertens 1998;11:978 982. 28. Kerestan SP: Enhanced hypertensive patient health care: performance of the B-D Assure blood pressure monitors. Personal communication, July 1998. 29. Personal communication, A&D, Jerry Wang. 30. Rocha JC, Rocha AT, Magossi AMG, Leao RW, Palu MJF, Moreira DC: Evaluation of the technique for taking blood pressure by health care workers in a university hospital (Abstract). Am J Hypertens 1988;11:66A. 31. Kay LE: Accuracy of blood pressure measurement in the family practice center. JABFP 1998;11:2528. 32. Veerman DP, van Montfrans GA, Wieling W: Effects of cuff inflation on self-recorded blood pressure. Lancet 1990;335:451 453. 33. Yarows SA: Home blood pressure monitoring in primary care. Blood Press Monit 1998;3(suppl 1):S11 S17. 34. Watson S, Wenzel RR, di Matteo C, Meier B, Luscher TF: Accuracy of a new wrist cuff oscillometric blood pressure device. Am J Hypertens 1998;11:1469 1474. 35. Personal communication, Omron Healthcare, Osamu Shirasaki, Director of Technology, September 2, 1998. 36. O Brien E, Mee F, Atkins N, Thomas M: Evaluation of three devices for self-measurement of blood pressure according to the revised British Hypertension Society Protocol: the Omron HEM-705CP, Philips HP5332, and Nissei DS-175. Blood Press Monitor 1996;1:55 61. 37. Eckert S, Gleichmann S, Gleichmann U: Blood pressure self-measurement in upper arm and in wrist for treatment of arterial hypertension. Zeitschrift für Kardiologie. 1996;85:109 111. 38. Izumi K: A comparison between blood pressure measurements taken with an Omron digital wrist blood pressure monitor (Model HEM-601) and those taken by the asculatory method. Data on file at Omron Healthcare, Inc. 39. Wambach G, Jacob R, Treeger A, Stimpel M: [How reliable is oscillometric blood pressure determination at the finger using the Omron device?] [German] Schweizerische Rundschau für Medizin Praxis 1992;81: 1258 1261. 40. Veerman DP, Lenders JW, Thien T, van Montfrans GA: LAM 100/Marshall F-88: accuracy and precision of a new device for discontinuous finger blood pressure measurement. J Hum Hypertens 1993;7:113 115. 41. Teshima M, Kuwajima I, Inukai M, Suzuki Y, Matsusita S, Kuramoto K: [Clinical evaluation of finger blood pressure measurement devices for home-use] [Japanese] Nippon Ronen Igakkai Zasshi Jpn J Geriatr 1993;30:54 58. 42. Iyriboz Y: Oscillometric finger blood pressure versus brachial auscultative blood pressure recording. J Fam Pract 1990;31:376 380. 43. Thijs L, Staessen JA, Celis H, de Gaudemaris R, Imai Y, Julius S, Falard R: Reference values for self-recorded blood pressure. A meta-analysis of summary data. Arch Intern Med 1998;158:481 488.