Automated Blood Pressure Determination During Exercise Test. Clinical Evaluation of a New Automated Device
Pietro Amedeo Modesti, M.D., Ph.D. Nazario Carrabba, M.D. Gian Franco Gensini, M.D. Francesco Bonechi, M.D. Tamara Taddei, M.D. and Pier Luigi Malfanti, M.D.
FLORENCE, ITALY
Abstract The accuracy and reproducibility of a new automatic device (P) specially designed for noninvasive blood pressure monitoring during the exercise stress test were evaluated in 50 consecutive subjects (34 normotensives and 16 hypertensives). Automatic measurements were compared with those taken by a sphygmomanometer (RR). A good agreement between systolic pressure values obtained by the two methods was found (RR 159±30 mmHg, P 158 ± 28 mmHg, mean difference = -1.53 ± 13 mmHg, p = 0. 166, ns). On the contrary the new device significantly underestimated diastolic pressure values (RR 89.3 ± 13 mmHg; P 84±13 mmHg, mean difference -5.37±9.3, p < 0.001). In conclusion the new device seems able to measure systolic but underestimates diastolic blood pressure both in hypertensives and in normotensives during the effort test. Introduction The assessment of blood pressure (BP) during the exercise stress test is routinely performed by noninvasive auscultatory technique.’-3 In recent years several automatic noninvasive blood pressure-monitoring devices have been introduced into clinical use4 but have often failed to accurately record blood pressure values during the effort test. ~’ In fact the devices employing the auscultatory technique (microphone) are disturbed by the ambient noise and by the From the Clinica Medica I,
University
of Florence, Florence
Italy
980
Downloaded from ang.sagepub.com at UNIV CALIFORNIA SAN DIEGO on August 26, 2015
981 noise produced under the cuff where the microphone is placed, whereas the devices employing oscillometry are disturbed by the upper limb motion of patients. A new, automatic noninvasive device (P) that employs both auscultatory (microphone) and oscillometric methods has been introduced. However, this device filters any noise different from the Korotkoff sounds by performing a spectral analysis of sounds perceived by its microphone during the inflation of the cuff so that the nonperiodical noises are refused. In this paper the accuracy and the reliability of this device in BP measurement during the effort test were evaluated by comparing automatic readings with those simultaneously obtained by a sphygmomanometer (RR), which in spite of its limitations, still represents the standard device employed routinely for clinical use.
Materials and Methods
Subjects Investigated Fifty subjects (36 men and 14 women) consecutively undergoing the exercise stress test (treadmill), with a mean age of 50.2 ~ 16.4 (range twenty-one to seventy years) and a mean body weight of71.7±17.5 kg (range 49-91), were recruited for the study. Subjects were divided into two groups (normotensives and hypertensives) according to their blood pressure values. Thirty-four were normotensives with a mean systolic blood pressure of 122 ±17 mmHg (range 100-135) and a mean diastolic pressure of 82 ± 10 mmHg (range 65-89). Sixteen subjects had diastolic blood pressure values above 90 mmHg on at least three different occasions (hypertensives group), with a mean systolic pressure of 168.5 f 6.2 mmHg (range 140180) and a diastolic pressure of 104. 5 ± 5 mmHg (range 100-110) . Blood Pressure Measurement The new device, which measures blood pressure by noninvasive auscultatory technique, is connected to the cuff by a pneumatic tube and to the microphone, placed under the cuff upon the brachial artery 4-5 cm over the elbow, by a cable. The sounds perceived by the acoustic transduction undergo spectral analysis so that Korotkoff sounds, which are regularly periodic and synchronous with the heart pulse, are separated from internal and external nonsynchronous noise. When blood pressure measurement is triggered, the hose starts inflating the cuff with an inflating velocity adjustable between 3 and 9 mmHg/sec. During inflation of the cuff the microphone continuously perceives the sounds generated in the brachial artery up to their disappearance occurring at the occlusion of the vessel. The instrument then filters by spectral analysis the graph of the intensity of the perceived sounds over the time. The first perceived sound validated by this procedure is taken as diastolic pressure (phase V of Korotkoff sounds) and the disappearance of the sounds is taken as systolic pressure (phase I of Korotkoff sounds). The correspondence between the two techniques was assessed at rest by performing three sequential BP measurements by the new device (P) at one upper limb and simultaneously by RR, according to the recommendation of the American Heart Associationg (AHA), at the opposite upper limb. The symmetry of blood pressure at both limbs was preliminarily checked and the side of instrument use (left or right) was randomly assigned. The mean and standard deviation (SD) of the three measurements were calculated for each method and for each patient.
Downloaded from ang.sagepub.com at UNIV CALIFORNIA SAN DIEGO on August 26, 2015
982
Patients
were
surements were
then exercised
simultaneously
treadmill according to the M-Bruce protocol, and meataken by the two methods at the end of the second minute of
by
a
each step. Editing Criteria and Statistical Analysis Values of diastolic pressure below 30 mmHg and systolic pressure below 50 mmHg were considered unreliable and the measurements were repeated. In addition, in the case that during exercise test the number of reliable values was less than 70 % , the recording was considered unsuccessful. Measurements taken by the two methods were compared by linear regression with 95 % confidence limits (estimate of mean y for given x), prediction limits (estimate of individual y for given x), and by Student’s t test for paired data for all the subjects investigated. Systematic differences between the two methods at rest and at each exercise step were assessed by analysis of variance (ANOVA). The blood pressure pattern during the whole test was compared by multivariate analysis of variance (MANOVA) . The blood pressure values of normotensives and
hypertensives
were
separately analyzed. Results
In 8 subjects, 6 normotensives and 2 hypertensives, the new device was unable to take reliable values and the subjects were removed from the study. During the test the microphone needed to be replaced in 5 subjects because the recorded values were unreliable. In particular 4.6 % of total systolic and 8 % of total diastolic measurements differed by more than 40 mmHg from values measured by sphygmomanometer. However, no immediately subsequent pressure determination differed by more than 10 mmHg from that performed by sphygmomanometer. No significant difference was observed for systolic pressure at Student’s t test for paired
FIG. 1.
Comparison
of concurrent manual and automated
systolic blood
pressure. R=correlation coefficient.
Downloaded from ang.sagepub.com at UNIV CALIFORNIA SAN DIEGO on August 26, 2015
983
(RR 159 9 30 mmHg, P 158 8 28 mmHg, mean difference= -1.53±13 mmHg, t= -1.39, p=0.166, ns). Systolic values measured by the two methods showed significant linear correlation (R=0.90, p < 0.001) (Fig. 1).
data
systolic pressure at rest did not show significant differences between the two techniques (-1.33 ~6.18 mmHg, p=0.834). During the exercise test the analysis of variance did not show significant differences between the two measurement techniques at any step and the overall mean difference between the two methods was not significant (Table I) (Fig. 3). The blood pressure pattern obtained during all the exercise tests by the two devices showed a good agreement (MANOVA test, F=0.39, p=0.857). In hypertensives the values of the systolic pressure recorded by the two techniques showed a significant difference at rest (p < 0.02), which disappeared during the effort test (Table II) (Fig. 4). However, the blood pressure pattern obtained during all the exercise tests by the two devices showed a good agreement (MANOVA F =0.16, p =0.977) . Diastolic pressure values were significantly underestimated by the new device at Student’s t test (RR 89 . 3 ± 13 mmHg; P 84 ± 13 mmHg, mean difference = - 5 . 37 ± 9 . 3 , t = - 6 . 75 , p < 0.001). For diastolic pressure a low correlation between the two methods was found (R=0.76, p < 0.001). In normotensives the diastolic pressure measured by the new device at rest and during the exercise test at each step was slightly but not significantly lower than that measured by the sphygmomanometer. However, the overall mean difference between the two methods showed a significant underestimation of the new device’s values (p < 0.001 at ANOVA test) (Table I) (Fig. 2). The diastolic BP pattern obtained by the two techniques during all the exercise tests was similar (MANOVA F=0.12, p=0.988). In normotensives the
TABLE I
TABLE II
Systolic and Diastolic Blood Pressure at Rest and at Different Levels of Exercise Recorded by the Two Techniques in 28 Normotensive Subjects
Systolic and Diastolic Blood Pressure at Rest and at Different Levels of Exercise Recorded by the Two Techniques in 14 Hypertensive Patients
Downloaded from ang.sagepub.com at UNIV CALIFORNIA SAN DIEGO on August 26, 2015
984
FIG. 2.
Comparison
of concurrent manual and automated diastolic blood pressure. R=correlation coefficient.
FIG. 3. Upper graph: Comparison of systolic blood pressure recorded in normotensives by Riva Rocci sphygmomanometer (---) and Paramed (-) at rest and during the effort test (F=0.184, p=0.669 at ANOVA
test).
I
Lower graph: Comparison of diastolic blood pressure recorded in normotensives by Rica Rocci sphygmomanometer (---) and Paramed (-) at rest and during the effort test (F =11.386, p < 0.001 at ANOVA
test).
Downloaded from ang.sagepub.com at UNIV CALIFORNIA SAN DIEGO on August 26, 2015
985
FIG. 4. Upper graph: Comparison of systolic blood pressure recorded in hypertensives by Riva Rocci sphygmomanometer (---) and P (-) at rest and during the effort test (F=2.46, p=0.119 at ANOVA test). Lower graph: Comparison of diastolic blood pressure recorded in hypertensives by Rica Rocci sphygmomanometer (---) and P (-) at rest and during the effort test (F = 8.16, p < 0.005 at ANOVA test).
In hypertensives the diastolic pressure recorded by the new device rest was significantly lower at the first, the second, and the third exercise step, but the correspondence improved at the last step (Table II) (Fig. 4). The overall mean difference showed a significant underestimation of the new device’s values (F = 8.158, p < 0.005 at ANOVA test), whereas the diastolic pattern obtained during all the exercise tests by the two devices was similar (MANOVA
F=1.07, p=0.380). Discussion
According to our findings, the new device significantly underestimated diastolic pressure both in hypertensives and in normotensives during the effort test when compared with auscultatory readings. This automated device records Korotkoff sounds while inflating the cuff, and because its sensitivity is higher than that of the human ear, it is likely that it is able to perceive sounds before the fifth phase occurs, ie, at lower BP values. During the effort test the agreement improved only at the last step, probably because the increase of intensity of the Korotkoff sounds at the peak of the stress test cancels the difference in the threshold of sensitivity between the new device and the human ear. Taking blood pressure readings while inflating the cuff differs from the recommendation of the AHAg and makes it impossible to take a simultaneous blood pressure measurement by a sphygmomanometer at
Downloaded from ang.sagepub.com at UNIV CALIFORNIA SAN DIEGO on August 26, 2015
986 the the
upper limb. To overcome this problem, we concomitantly recorded blood pressure at opposite upper limb. This method has been employed both for noninvasive’ and invasive methods4.6,7.9 even though it is associated with the hypothetical limitation of variability of pressure at the opposite upper limb during exercise. In 16 % of patients the new device was unable to measure systolic BP even at rest. In those patients the disagreement between the two methods was still present during the test. In fact the placement of the microphone over the brachial artery under the cuff is critical, since when a good agreement was observed at rest, the same agreement was maintained during the performance of the test. During the exercise test only 4.6% of total systolic and 8% of total diastolic measurements taken by the new device were unreliable. However, the prompt repeating of automated measurements provided correct values. Systolic pressure is a very important element during the effort test and its values have to be carefully taken because its undue increase represents a criterion for the interruption of the test. However, the measurement of arterial pressure during an effort test is often difficult,2 particularly because of environmental noise and motion artifacts. The availability of a reliable automated device to measure BP during the stress tests could be useful for obtaining objective BP data. Conventional automated blood pressure recorders often fail to take a measurement during the effort test.4,5,7 The new device does not seem to be influenced by external noise and it maintains the same correspondence of systolic values even during the test. same
Conclusion Automatic pressure measurements taken by the new device during the exercise test may provide a useful auxiliary help to the physician while monitoring the patient. In fact even if a correct measurement of diastolic values is often lacking, the device is able to follow the pattern of systolic pressure, without being affected by external noise.
Acknowledgments The authors ments tested
are
most
grateful to ESAOTE Biomedica, Firenze, 9350, Technology Inc. USA).
for
supplying
the instru-
(Paramed mod.
Pietro Amedeo Modesti, M ., Ph. D. D . Clinica Medica I , Viale Morgagni 85 50134 - Florence, Italy
Downloaded from ang.sagepub.com at UNIV CALIFORNIA SAN DIEGO on August 26, 2015
987
References 1.
Irving JB, Bruce RA, DeRouen TA: Variations in and
systolic pressure during maximal exer(treadmill) testing. Am J Cardiol 39:841-848,
significance
of
cise 1977.
Petrie A: Clinical experience of ultrasound sphygmomanometer. Br Heart J 7:804-807, 1975. 3. Berliner K: The accuracy of blood pressure determination: A comparison of direct and indirect measurements. Cardiologia 37:118-128, 1960. 4. White WB, Lund-Johansen P, Omvik P: Assessment of four ambulatory blood pressure monitors and measurements by clinicians versus intraarterial blood pressure at rest and during exercise. Am J Cardiol 65:60-66, 1990. 5. Modesti PA, Gensini GF, Conti C, et al: Clinical evaluation of an automatic blood pressure monitoring device. J
2.
George CF, Lewis PJ, with
use
Clin Hypertens 3:631-644, 1987. 6. Takahashi H, Yoshimura, Nishimura M, et al: Measurement of digital arterial pressure in patients with essential hypertension. Jpn Circ J 54:221-230, 1990. 7. Idema RN, van den Meiracker AH, Imholz BP, et al: Comparison of Finapres non-invasive beat-to-beat finger blood pressure with intrabrachial artery pressure during and after bicycle ergometry. J Hypertens 7 (Suppl):S58-59, 1989. 8. Frohlich ED, Grim C, Labarthe DR, et al: Recommendations for human blood pressure determination by sphygmomanometers. Circulation 77:502A-514A, 1988. 9. Gravlee GP, Brockschmidt JK: Accuracy of four indirect methods of blood pressure measurement, with hemodynamic correlations. J Clin Monit 6:284-298, 1990.
Downloaded from ang.sagepub.com at UNIV CALIFORNIA SAN DIEGO on August 26, 2015
Pietro Amedeo Modesti, M.D., Ph.D. Nazario Carrabba, M.D. Gian Franco Gensini, M.D. Francesco Bonechi, M.D. Tamara Taddei, M.D. and Pier Luigi Malfanti, M.D.
FLORENCE, ITALY
Abstract The accuracy and reproducibility of a new automatic device (P) specially designed for noninvasive blood pressure monitoring during the exercise stress test were evaluated in 50 consecutive subjects (34 normotensives and 16 hypertensives). Automatic measurements were compared with those taken by a sphygmomanometer (RR). A good agreement between systolic pressure values obtained by the two methods was found (RR 159±30 mmHg, P 158 ± 28 mmHg, mean difference = -1.53 ± 13 mmHg, p = 0. 166, ns). On the contrary the new device significantly underestimated diastolic pressure values (RR 89.3 ± 13 mmHg; P 84±13 mmHg, mean difference -5.37±9.3, p < 0.001). In conclusion the new device seems able to measure systolic but underestimates diastolic blood pressure both in hypertensives and in normotensives during the effort test. Introduction The assessment of blood pressure (BP) during the exercise stress test is routinely performed by noninvasive auscultatory technique.’-3 In recent years several automatic noninvasive blood pressure-monitoring devices have been introduced into clinical use4 but have often failed to accurately record blood pressure values during the effort test. ~’ In fact the devices employing the auscultatory technique (microphone) are disturbed by the ambient noise and by the From the Clinica Medica I,
University
of Florence, Florence
Italy
980
Downloaded from ang.sagepub.com at UNIV CALIFORNIA SAN DIEGO on August 26, 2015
981 noise produced under the cuff where the microphone is placed, whereas the devices employing oscillometry are disturbed by the upper limb motion of patients. A new, automatic noninvasive device (P) that employs both auscultatory (microphone) and oscillometric methods has been introduced. However, this device filters any noise different from the Korotkoff sounds by performing a spectral analysis of sounds perceived by its microphone during the inflation of the cuff so that the nonperiodical noises are refused. In this paper the accuracy and the reliability of this device in BP measurement during the effort test were evaluated by comparing automatic readings with those simultaneously obtained by a sphygmomanometer (RR), which in spite of its limitations, still represents the standard device employed routinely for clinical use.
Materials and Methods
Subjects Investigated Fifty subjects (36 men and 14 women) consecutively undergoing the exercise stress test (treadmill), with a mean age of 50.2 ~ 16.4 (range twenty-one to seventy years) and a mean body weight of71.7±17.5 kg (range 49-91), were recruited for the study. Subjects were divided into two groups (normotensives and hypertensives) according to their blood pressure values. Thirty-four were normotensives with a mean systolic blood pressure of 122 ±17 mmHg (range 100-135) and a mean diastolic pressure of 82 ± 10 mmHg (range 65-89). Sixteen subjects had diastolic blood pressure values above 90 mmHg on at least three different occasions (hypertensives group), with a mean systolic pressure of 168.5 f 6.2 mmHg (range 140180) and a diastolic pressure of 104. 5 ± 5 mmHg (range 100-110) . Blood Pressure Measurement The new device, which measures blood pressure by noninvasive auscultatory technique, is connected to the cuff by a pneumatic tube and to the microphone, placed under the cuff upon the brachial artery 4-5 cm over the elbow, by a cable. The sounds perceived by the acoustic transduction undergo spectral analysis so that Korotkoff sounds, which are regularly periodic and synchronous with the heart pulse, are separated from internal and external nonsynchronous noise. When blood pressure measurement is triggered, the hose starts inflating the cuff with an inflating velocity adjustable between 3 and 9 mmHg/sec. During inflation of the cuff the microphone continuously perceives the sounds generated in the brachial artery up to their disappearance occurring at the occlusion of the vessel. The instrument then filters by spectral analysis the graph of the intensity of the perceived sounds over the time. The first perceived sound validated by this procedure is taken as diastolic pressure (phase V of Korotkoff sounds) and the disappearance of the sounds is taken as systolic pressure (phase I of Korotkoff sounds). The correspondence between the two techniques was assessed at rest by performing three sequential BP measurements by the new device (P) at one upper limb and simultaneously by RR, according to the recommendation of the American Heart Associationg (AHA), at the opposite upper limb. The symmetry of blood pressure at both limbs was preliminarily checked and the side of instrument use (left or right) was randomly assigned. The mean and standard deviation (SD) of the three measurements were calculated for each method and for each patient.
Downloaded from ang.sagepub.com at UNIV CALIFORNIA SAN DIEGO on August 26, 2015
982
Patients
were
surements were
then exercised
simultaneously
treadmill according to the M-Bruce protocol, and meataken by the two methods at the end of the second minute of
by
a
each step. Editing Criteria and Statistical Analysis Values of diastolic pressure below 30 mmHg and systolic pressure below 50 mmHg were considered unreliable and the measurements were repeated. In addition, in the case that during exercise test the number of reliable values was less than 70 % , the recording was considered unsuccessful. Measurements taken by the two methods were compared by linear regression with 95 % confidence limits (estimate of mean y for given x), prediction limits (estimate of individual y for given x), and by Student’s t test for paired data for all the subjects investigated. Systematic differences between the two methods at rest and at each exercise step were assessed by analysis of variance (ANOVA). The blood pressure pattern during the whole test was compared by multivariate analysis of variance (MANOVA) . The blood pressure values of normotensives and
hypertensives
were
separately analyzed. Results
In 8 subjects, 6 normotensives and 2 hypertensives, the new device was unable to take reliable values and the subjects were removed from the study. During the test the microphone needed to be replaced in 5 subjects because the recorded values were unreliable. In particular 4.6 % of total systolic and 8 % of total diastolic measurements differed by more than 40 mmHg from values measured by sphygmomanometer. However, no immediately subsequent pressure determination differed by more than 10 mmHg from that performed by sphygmomanometer. No significant difference was observed for systolic pressure at Student’s t test for paired
FIG. 1.
Comparison
of concurrent manual and automated
systolic blood
pressure. R=correlation coefficient.
Downloaded from ang.sagepub.com at UNIV CALIFORNIA SAN DIEGO on August 26, 2015
983
(RR 159 9 30 mmHg, P 158 8 28 mmHg, mean difference= -1.53±13 mmHg, t= -1.39, p=0.166, ns). Systolic values measured by the two methods showed significant linear correlation (R=0.90, p < 0.001) (Fig. 1).
data
systolic pressure at rest did not show significant differences between the two techniques (-1.33 ~6.18 mmHg, p=0.834). During the exercise test the analysis of variance did not show significant differences between the two measurement techniques at any step and the overall mean difference between the two methods was not significant (Table I) (Fig. 3). The blood pressure pattern obtained during all the exercise tests by the two devices showed a good agreement (MANOVA test, F=0.39, p=0.857). In hypertensives the values of the systolic pressure recorded by the two techniques showed a significant difference at rest (p < 0.02), which disappeared during the effort test (Table II) (Fig. 4). However, the blood pressure pattern obtained during all the exercise tests by the two devices showed a good agreement (MANOVA F =0.16, p =0.977) . Diastolic pressure values were significantly underestimated by the new device at Student’s t test (RR 89 . 3 ± 13 mmHg; P 84 ± 13 mmHg, mean difference = - 5 . 37 ± 9 . 3 , t = - 6 . 75 , p < 0.001). For diastolic pressure a low correlation between the two methods was found (R=0.76, p < 0.001). In normotensives the diastolic pressure measured by the new device at rest and during the exercise test at each step was slightly but not significantly lower than that measured by the sphygmomanometer. However, the overall mean difference between the two methods showed a significant underestimation of the new device’s values (p < 0.001 at ANOVA test) (Table I) (Fig. 2). The diastolic BP pattern obtained by the two techniques during all the exercise tests was similar (MANOVA F=0.12, p=0.988). In normotensives the
TABLE I
TABLE II
Systolic and Diastolic Blood Pressure at Rest and at Different Levels of Exercise Recorded by the Two Techniques in 28 Normotensive Subjects
Systolic and Diastolic Blood Pressure at Rest and at Different Levels of Exercise Recorded by the Two Techniques in 14 Hypertensive Patients
Downloaded from ang.sagepub.com at UNIV CALIFORNIA SAN DIEGO on August 26, 2015
984
FIG. 2.
Comparison
of concurrent manual and automated diastolic blood pressure. R=correlation coefficient.
FIG. 3. Upper graph: Comparison of systolic blood pressure recorded in normotensives by Riva Rocci sphygmomanometer (---) and Paramed (-) at rest and during the effort test (F=0.184, p=0.669 at ANOVA
test).
I
Lower graph: Comparison of diastolic blood pressure recorded in normotensives by Rica Rocci sphygmomanometer (---) and Paramed (-) at rest and during the effort test (F =11.386, p < 0.001 at ANOVA
test).
Downloaded from ang.sagepub.com at UNIV CALIFORNIA SAN DIEGO on August 26, 2015
985
FIG. 4. Upper graph: Comparison of systolic blood pressure recorded in hypertensives by Riva Rocci sphygmomanometer (---) and P (-) at rest and during the effort test (F=2.46, p=0.119 at ANOVA test). Lower graph: Comparison of diastolic blood pressure recorded in hypertensives by Rica Rocci sphygmomanometer (---) and P (-) at rest and during the effort test (F = 8.16, p < 0.005 at ANOVA test).
In hypertensives the diastolic pressure recorded by the new device rest was significantly lower at the first, the second, and the third exercise step, but the correspondence improved at the last step (Table II) (Fig. 4). The overall mean difference showed a significant underestimation of the new device’s values (F = 8.158, p < 0.005 at ANOVA test), whereas the diastolic pattern obtained during all the exercise tests by the two devices was similar (MANOVA
F=1.07, p=0.380). Discussion
According to our findings, the new device significantly underestimated diastolic pressure both in hypertensives and in normotensives during the effort test when compared with auscultatory readings. This automated device records Korotkoff sounds while inflating the cuff, and because its sensitivity is higher than that of the human ear, it is likely that it is able to perceive sounds before the fifth phase occurs, ie, at lower BP values. During the effort test the agreement improved only at the last step, probably because the increase of intensity of the Korotkoff sounds at the peak of the stress test cancels the difference in the threshold of sensitivity between the new device and the human ear. Taking blood pressure readings while inflating the cuff differs from the recommendation of the AHAg and makes it impossible to take a simultaneous blood pressure measurement by a sphygmomanometer at
Downloaded from ang.sagepub.com at UNIV CALIFORNIA SAN DIEGO on August 26, 2015
986 the the
upper limb. To overcome this problem, we concomitantly recorded blood pressure at opposite upper limb. This method has been employed both for noninvasive’ and invasive methods4.6,7.9 even though it is associated with the hypothetical limitation of variability of pressure at the opposite upper limb during exercise. In 16 % of patients the new device was unable to measure systolic BP even at rest. In those patients the disagreement between the two methods was still present during the test. In fact the placement of the microphone over the brachial artery under the cuff is critical, since when a good agreement was observed at rest, the same agreement was maintained during the performance of the test. During the exercise test only 4.6% of total systolic and 8% of total diastolic measurements taken by the new device were unreliable. However, the prompt repeating of automated measurements provided correct values. Systolic pressure is a very important element during the effort test and its values have to be carefully taken because its undue increase represents a criterion for the interruption of the test. However, the measurement of arterial pressure during an effort test is often difficult,2 particularly because of environmental noise and motion artifacts. The availability of a reliable automated device to measure BP during the stress tests could be useful for obtaining objective BP data. Conventional automated blood pressure recorders often fail to take a measurement during the effort test.4,5,7 The new device does not seem to be influenced by external noise and it maintains the same correspondence of systolic values even during the test. same
Conclusion Automatic pressure measurements taken by the new device during the exercise test may provide a useful auxiliary help to the physician while monitoring the patient. In fact even if a correct measurement of diastolic values is often lacking, the device is able to follow the pattern of systolic pressure, without being affected by external noise.
Acknowledgments The authors ments tested
are
most
grateful to ESAOTE Biomedica, Firenze, 9350, Technology Inc. USA).
for
supplying
the instru-
(Paramed mod.
Pietro Amedeo Modesti, M ., Ph. D. D . Clinica Medica I , Viale Morgagni 85 50134 - Florence, Italy
Downloaded from ang.sagepub.com at UNIV CALIFORNIA SAN DIEGO on August 26, 2015
987
References 1.
Irving JB, Bruce RA, DeRouen TA: Variations in and
systolic pressure during maximal exer(treadmill) testing. Am J Cardiol 39:841-848,
significance
of
cise 1977.
Petrie A: Clinical experience of ultrasound sphygmomanometer. Br Heart J 7:804-807, 1975. 3. Berliner K: The accuracy of blood pressure determination: A comparison of direct and indirect measurements. Cardiologia 37:118-128, 1960. 4. White WB, Lund-Johansen P, Omvik P: Assessment of four ambulatory blood pressure monitors and measurements by clinicians versus intraarterial blood pressure at rest and during exercise. Am J Cardiol 65:60-66, 1990. 5. Modesti PA, Gensini GF, Conti C, et al: Clinical evaluation of an automatic blood pressure monitoring device. J
2.
George CF, Lewis PJ, with
use
Clin Hypertens 3:631-644, 1987. 6. Takahashi H, Yoshimura, Nishimura M, et al: Measurement of digital arterial pressure in patients with essential hypertension. Jpn Circ J 54:221-230, 1990. 7. Idema RN, van den Meiracker AH, Imholz BP, et al: Comparison of Finapres non-invasive beat-to-beat finger blood pressure with intrabrachial artery pressure during and after bicycle ergometry. J Hypertens 7 (Suppl):S58-59, 1989. 8. Frohlich ED, Grim C, Labarthe DR, et al: Recommendations for human blood pressure determination by sphygmomanometers. Circulation 77:502A-514A, 1988. 9. Gravlee GP, Brockschmidt JK: Accuracy of four indirect methods of blood pressure measurement, with hemodynamic correlations. J Clin Monit 6:284-298, 1990.
Downloaded from ang.sagepub.com at UNIV CALIFORNIA SAN DIEGO on August 26, 2015
