Radial artery tonometry: moderately accurate but unpredictable technique of continuous non-invasive arterial pressure measurement
7
0
0
Texte intégral
(2) 406. British Journal of Anaesthesia. paediatric patients was observed [6]. An accurate technique of CNAP measurement, however, could replace invasive arterial pressure monitoring in some patients, avoiding the risks of arterial cannulation. In addition, arterial cannulation can be difficult in the awake state, and CNAP monitoring may be a temporary alternative, allowing cannulation later under the more favourable condition of general anaesthesia. The purpose of the study was to evaluate radial artery tonometry in patients with concomitant cardiac, vascular, pulmonary and renal diseases undergoing non-thoracic surgical procedures.. Patients and methods The study was approved by the Institutional Ethics Committee. After obtaining informed consent, we studied patients undergoing abdominal or transurethral surgical procedures in the supine or lithotomy position, requiring invasive arterial pressure monitoring with no history of recent radial artery cannulation. The anaesthetic technique was chosen independent of the investigation. Premedication comprised oral midazolam and patients were placed in the supine position on an operating table. A circulating water mattress and warmed i.v. infusions were used to maintain normothermia. The arms were placed on armholders at the level of the heart, abducted approximately 45–70 from the body. Arterial pressure was measured oscillometrically on both arms, using an adult size, inflatable upper arm cuff and a multifunctional monitoring instrument (BP-508, Colin Medical Instrumentations, Komaki, Japan). Three patients with an inter-arm mean arterial pressure difference exceeding 10 mm Hg (average of three measurements) were excluded. Patient characteristics (19 males and three females), inter-arm oscillometric differences and concomitant diseases are summarized in table 1. A radial artery cannula (20-gauge, 32 mm) was inserted during local anaesthesia and connected to low compliance tubing (diameter 1.0–1.5 mm, length 165 cm, two stopcocks). A disposable pressure transducer (Deltran, Utah Medical, Midvale, UT, USA) was used. This set was tested previously in Table 1 Patient characteristics (n : 22) (mean (SD) [range]) Age (yr) 66.3 [25.5–91.1] Weight (kg) 74.0 (14.3) [42.5–91.0] Height (cm) 172 (7) [155–182] Inter-arm pressure difference (mm Hg) Systolic 0.1 (6.9) [94.3–14.3] Diastolic 1.4 (6.4) [96.7–15.7)] Mean 3.0 (6.2) [99.6–10.0] Concomitant diseases (n) Coronary artery disease 11 Arterial hypertension 9 Chronic heart failure 6 Cardiac arrhythmias 3 Valvular heart disease 1 Peripheral vascular disease 5 Cerebrovascular disease 4 Diabetes mellitus 4 Chronic lung disease 5 Renal insufficiency 5. vitro with a pressure waveform generator. Almost identical values for mean and trough pressure were found, and peak pressure was overestimated by a mean of 2.0 (SD 0.8) mm Hg, compared with values obtained by a micromanometer pressure transducer [7]. The pressure transducer was connected to the pressure module (with a low-pass filter for frequencies greater than 25 Hz) of the Colin instrument. The level of the right atrium was used for zero calibration. The upper arm cuff and the housing with a multielement array sensor (SA-250, Colin, Komaki, Japan), attached above the radial artery, were placed on the contralateral arm. The dimensions of the sensor housing were 7.04.03.7 cm, with weight 90 g. The wrist was placed in a brace in an extended position and the brace and sensor housing were wrapped, as recommended by the manufacturer. The intermittent oscillometric measurements served as the reference (occlusive calibration) and defined the range and values for tonometric CNAP. Automatic recalibration occurred independently of the preset interval whenever the device registered distortion of the CNAP curve. The instrument displayed on the screen: digital oscillometric pressure values, selected time interval and time elapsed since the previous calibration; invasive and CNAP curves on the same scale; and synchronously, the updated invasive and tonometric digital values for systolic, diastolic and mean arterial pressures. One of the 30 elements of the sensor transducer identified automatically maximal pulse amplitude signals from the flattened radial artery. A small graphical display (tonogram) on the screen indicated the position of the “active transducer” as a vertical line and as a bar. Within given limits, automatic repositioning of the active element–transducer and optimal pulse amplitude identification occurred, otherwise manual repositioning was required. Another graphical display and digital value indicated so-called signal strength. Signal strength is calculated automatically by the device as the ratio between the peak-to-peak amplitudes of tonometric pressure to the amplitudes of cuff oscillations. According to the manufacturer, signal strength should be 50–100 amplitude units. Inadequate sensor placement and a signal strength less than 50 amplitude units required repositioning of the sensor. The pressure applied on the wrist by the sensor (hold-down pressure), controlled through a microprocessor and set by a pneumatic bladder within the housing, was displayed on the screen. Initial signal strength and hold-down pressure values were recorded in each patient, and controlled throughout the procedure. Comparative measurements were started before induction of anaesthesia and lasted until the patient was prepared to leave the operating room. Oscillometric calibration was preset at 5-min intervals during induction of anaesthesia and at 10-min intervals for the rest of the procedure. Invasive and tonometric CNAP values were documented every 30–60 s during induction of anaesthesia and every 3–5 min for the remainder of the procedure. The built-in printer automatically and continually recorded the time of day and a pair of oscillometric.
(3) Radial artery tonometry in high-risk patients and invasive pressure values or a pair of invasive and tonometric CNAP values. All values were later checked and incomplete or artefactual pairs (collected during blood aspiration, flushing, repositioning of the sensor or unstable signal strength) were excluded from analysis. As the waveform recording was not provided by this instrument software, both CNAP and invasive pressure curves were printed continually on a multichannel paper recorder (Servomed, Hellige, Freiburg, Germany). The oscillometric signal and scale could not be recorded, and the CNAP trace was centred manually within the channel and the amplitude adjusted to that of invasive pressure. The time periods of induction of anaesthesia and surgical procedure, actual time interval between oscillometric calibrations and actual frequency of CNAP collection were recorded for each patient. Invasive pressure was considered the reference. Oscillometric and corresponding invasive pressure pairs and, separately, tonometric CNAP and corresponding invasive pressure pairs were analysed. Pressure differences were calculated for each pair of systolic, diastolic and mean values. Mean (SD) pressure differences (prediction error or bias) were derived for oscillometry and tonometry, cumulatively for all data and individually for each patient. Absolute error (precision) was calculated as the sum of squared pressure differences, divided by the number of measurements, cumulatively for all data and individually for each patient [8]. Percentage error was derived from (mean pressure difference/ mean of average pressure) 100. Oscillometric and tonometric errors were compared by the Student’s t test and P 0.05 was considered statistically significant. To evaluate the correlation between the measured and calculated variables, linear regression analysis, with coefficient of correlation (r) and 95 % confidence intervals for the single observations, was used. The average pressure of each pair and the mean of the average were calculated for oscillometric and invasive, and for tonometric and invasive pairs. Each tonometric prediction error was plotted against the average pressure value of the two measurements, and the mean error (bias) and limits of agreement (<2 SD) were drawn [9]. Each patient’s individual prediction error of tonometric and oscillometric systolic, diastolic and mean arterial pressure was defined arbitrarily as good (5 (8) mm Hg), acceptable (5 (8) errors 10 (12) mm Hg) or poor (10 (12) mm Hg). Individual mean absolute errors of 10, 11–15 and 15 mm Hg were arbitrarily classified as good, acceptable and poor, respectively. Bias and precision of systolic, diastolic and mean arterial pressures were combined to assess oscillometry and, separately, tonometry in each patient, according to an arbitrary score (two errors for three pressures: 2 : good, 1 : acceptable and 0 : poor error). Good monitoring required a score of 9–12 points, acceptable monitoring 5–8 points, and 0–4 points indicated a poor function of tonometric CNAP. A pressure difference between two consecutive measurements (trend of changes) was derived for the invasive and CNAP methods during induction of anaesthesia. The critical difference. 407 between the two trends was arbitrarily set at 10 mm Hg. Data are presented as mean (SD) unless otherwise stated.. Results The mean period of measurement was 153 (87) min and the mean duration of surgery 109 (75) min. We studied five patients during spontaneous breathing, using regional anaesthesia in two patients and general anaesthesia in three. General anaesthesia, neuromuscular block and controlled ventilation were used in 17 patients, in seven patients combined with a continuous extradural block. Single or multiple i.v. injections of ephedrine, nitroglycerin, or both, were given to 12 patients, one patient received a dopamine infusion of 1–3 g kg91 min91 for 2.5 h, and in nine patients vasoactive drugs were not given. Intraoperative blood loss was estimated as less than 10 % of circulating blood volume in 13 patients, less than 20 % in two patients, less than 50 % in three patients and exceeding 50 % in four patients. Blood volume was substituted, as required, with crystalloid and colloid solutions, and blood transfusions. According to the tonogram, and axial and signal strength indicators, correct sensor positioning was obtained initially in all patients. Initial values of signal strength and hold-down pressure were 96.5 (21.8) amplitude units and 78.0 (32.7) mm Hg, respectively. Protection of the sensor position, as a result of the abducted arm and the weight and volume of the sensor housing itself, was found to be difficult on occasion. In several patients repositioning of the sensor was required during the study. In one patient tonometric CNAP ceased to function, the warnings and repeated oscillometric measurements indicated a function failure, and then CNAP monitoring returned automatically. The function of the device was not disturbed by the intraoperative use of electrocautery. The CNAP curve looked acceptable, compared with the simultaneous invasive pressure curve, during normal heart rate and cardiac arrhythmia. We analysed a total of 481 pairs of oscillometric and invasive pressure (21.9 (12.6) pairs per patient), and 1375 pairs of tonometric and invasive pressure (62.5 (28.4) pairs per patient). During induction of anaesthesia (30.7 (18) min) and for the remainder of the procedure, oscillometric calibration was performed every 6.1 (3.1) min and 8.5 (2.5) min, respectively, and tonometric CNAP data were collected every 1.0 (0.5) min and 4.7 (2.0) min, respectively. Pressure differences and dispersion of errors showed a normal distribution. Cumulatively (table 2), oscillometric measurements were more accurate than tonometric CNAP. Absolute errors of CNAP were significantly larger than corresponding absolute errors of oscillometry. Average CNAP errors and linear regression analysis (r, mean prediction line) of the pooled data showed small discrepancies and a significant correlation with invasive pressure values. The confidence interval range, however, exceeded 50 mm Hg for systolic and 30 mm Hg for diastolic and mean (fig. 1) arterial pressures. The limits of agreement (2 SD) between.
(4) 408. British Journal of Anaesthesia Table 2 Cumulative analysis of non-invasive vs invasive arterial pressure values (mean (SD)). *** P 0.001 vs corresponding oscillometric errors Oscillometry (n : 481). Correlation (r) Prediction error (bias) (mm Hg) Average pressure (mm Hg) Percentage error of average (%) Absolute error (precision) (mm Hg). Tonometry (CNAP, n = 1375). Systolic. Diastolic. Mean. Systolic. Diastolic. Mean. 0.83 92.2 (12.8) 122.4 (20.4) 91.8 (10.4) 7.6 (5.1). 0.90 ;8.8 (5.1) ;66.2 (7.8) ;13.3 (7.8) 4.7 (3.1). 0.92 ;4.4 (6.2) 85.2 (14.3) ;5.2 (7.2) 2.6 (1.9). 0.80 95.8 (14.2) 121.0 (20.4) 94.8 (11.8) 15.2*** (4.7). 0.77 ;7.2 (8.3) 64.2 (11.5) ;11.1 (12.9) 10.9*** (3.0). 0.84 ;3.9 (8.8) 83.7 (14.9) ;4.6 (10.6) 9.4*** (3.4). Figure 1 Linear regression analysis with scatterplot for mean arterial pressure (MAP) of the invasive vs the CNAP method (n : 1375, r : 0.84). Mean prediction line and 95 % confidence intervals for the single observations, with a range of 33 mm Hg, are shown.. Figure 2 Scatterplot of pressure differences and average pressure for mean arterial pressure (MAP) of the invasive vs the CNAP method (n : 1375). Mean prediction error (bias) and limits of agreement (2 SD : 16.5 mm Hg) between the two methods are shown, according to Bland and Altman [9].. the invasive and CNAP methods exceeded 54 mm Hg for systolic and 30 mm Hg for diastolic and mean (fig. 2) arterial pressures. According to individual bias and precision of oscillometry, 14 (11 %) of 132 (two criteria, three. pressures, 22 patients) values were in the poor category. Globally, intermittent non-invasive monitoring was assessed as good in 16 and as acceptable in six patients. According to each patient’s individual bias and precision of tonometry, 50 % of systolic values showed poor accuracy, diastolic values were predominantly in the acceptable category and the majority of the mean CNAP values showed good agreement with invasive measurements (table 3). CNAP monitoring was assessed as good in four patients, acceptable in 14 and poor in four of 22 patients. These four patients with inadequate tonometric accuracy could not be distinguished from the other 18 patients in terms of concomitant diseases, inter-arm oscillometric pressure difference (in one case it was among the lowest of all patients), bias and precision of oscillometry, course of the procedure (blood loss, vasoactive drugs) and sensor performance (signal strength, hold-down pressure). Individual inter-arm pressure difference did not correlate with individual oscillometric (r : 0.13) or tonometric (r : 0.25) prediction errors. A strong correlation between oscillometric and tonometric prediction errors (r : 0.86, fig. 3), and a moderate correlation between oscillometric and tonometric absolute errors (r : 0.48) were found, but the confidence interval range for both analyses exceeded 15 mm Hg. The correlation of trends in mean arterial pressure changes with the invasive and CNAP methods (n : 616, 28 (16) pairs per patient) during induction of anaesthesia showed a confidence interval range within 20 mm Hg (fig. 4). The difference between the two trends exceeded 10 mm Hg in 47 (7.6 %) measurements for systolic pressure in 16 patients, in 14 (2.3 %) measurements for diastolic pressure in seven patients and in 27 (4.4 %) measurements for mean arterial pressure in 12 patients. These discrepancies occurred predominantly in one of 22 patients.. Discussion Comparison of non-invasive and invasive arterial pressure methods results inevitably in disagreement [10]. Measurements at different arterial sites are confounded by differences in velocity of pressure wave propagation, unequal impedance, reflections and resonances, in addition to differences in blood.
(5) Radial artery tonometry in high-risk patients. 409. Table 3 Accuracy of tonometric CNAP according to each patient’s individual prediction error (bias) and individual absolute error (precision). *Prediction error of 5 (8) mm Hg was arbitrarily defined as good, 5 (8) error 10 (12) mm Hg as acceptable and 10 (12) mm Hg as poor. Absolute errors of 10, 11–15 and 15 mm Hg were defined correspondingly. Data are number of patients (%) for each category (total : six criteria in 22 patients : 132 (100 %)) Prediction error (bias). Absolute error (precision). Definition*. Systolic. Diastolic. Mean. Systolic. Diastolic. Mean. Total. Good Acceptable Poor Patients. 1 (5) 9 (41) 12 (54) 22 (100). 2 (9) 16 (73) 4 (18) 22 (100). 14 (64) 5 (23) 3 (13) 22 (100). 3 (14) 8 (36) 11 (50) 22 (100). 8 (36) 11 (50) 3 (14) 22 (100). 16 (72) 5 (23) 1 (5) 22 (100). 44 (33) 54 (41) 34 (26) 132 (100). Figure 3 Linear regression analysis with scatterplot of individual oscillometric mean prediction error vs individual tonometric (CNAP) mean prediction error for systolic, diastolic and mean arterial pressures (n : 66, r : 0.86). Mean prediction line and 95 % confidence intervals for single observations, with a range of 17 mm Hg, are shown.. flow distribution within the arterial system [11]. Furthermore, the basic principles of non-invasive methods share no similarities with, and only superficially resemble, the invasive method [1–6, 10, 12–14]. Although the intra-arterial measurement should theoretically be considered the “gold standard”, its precision is dependent on a variety of technical factors [3–7, 10–17]. Also, numerous, poorly controlled patient variables, external influences and cardiocirculatory changes during a dynamic period of anaesthesia and surgery further accentuate the difference between the methods of pressure measurement [3–6, 10–24]. As a rule, pressure values from different methods over time show bidirectional variations in one subject and even more variation when the data are summarized for a group of patients. Different principles of measurement at different arterial sites, and different types of capturing, speed and interpretation of “pressure signals” explain the absence of exclusive or even predominant pressure over- or underestimation by a non-invasive method. The present study indicates that, compared with contralateral invasive radial artery pressure: intermittent oscillometry provides. Figure 4 Linear regression analysis with scatterplot of trends in mean arterial pressure (MAP) changes (difference between consecutive measurements) for the invasive and CNAP methods during induction of anaesthesia (n : 616, r : 0.66). Mean prediction line and 95 % confidence intervals for the single observations, with a range within 20 mm Hg, are shown. Data in the B and C quadrants indicate the changes in the two methods in the same direction, and data in the A and D quadrants indicate changes in the opposite direction. In 27 instances (4.4 %), predominantly in one of 22 patients, the difference between the two trends exceeded 10 mm Hg.. accurate arterial pressure monitoring; CNAP measurements by radial artery tonometry offer a reliable trend indicator of pressure changes during induction of anaesthesia; radial artery tonometry may be considered a valid alternative to invasive pressure measurements, if arterial cannulation is difficult in an awake patient; and accuracy of absolute CNAP values is only moderate and unpredictable, thus it should not replace invasive monitoring in high-risk patients during major surgical procedures. Other CNAP methods (finger photoplethysmomanometry, dual photometry and pulse wave-delay, brachial artery displacement) have been evaluated previously during anaesthetic and surgical procedures [23–29]. Using various criteria, each method has been shown to be insufficiently accurate for most subjects or for most of the time, or both, and none could be recommended as an alternative to invasive pressure monitoring. Radial artery tonometry had promise as an accurate and simple method of CNAP.
(6) 410 recording, with narrow limits of agreement and high precision compared with simultaneous invasive measurements [1]. A high level of accuracy was obtained during controlled hypotension but invasive pressure was measured in some cases with a resonance overshoot eliminator, without explaining the reasons and criteria for its use [4]. Such a device has not been tested adequately under clinical conditions [7, 16, 17], and thus cannot be considered a requirement for high-fidelity reproduction of invasive arterial pressure. More recently, Siegel, Brock-Utne and Brodsky analysed a large number of invasive and tonometric CNAP pairs, and reported discrepancies between invasive and CNAP readings [5]. These authors concluded that tonometric CNAP is unable to detect changes in arterial pressure more rapidly than intermittent oscillometry, as accuracy of the tonometric sensor is limited. For this investigation, a mixed group of high-risk surgical patients, undergoing various non-thoracic surgical procedures, was selected for testing tonometric CNAP. Both non-invasive methods, oscillometry and tonometry, on average underestimated systolic and overestimated diastolic and mean arterial pressures. Compared with intermittent oscillometry, radial artery tonometry exhibited inferior accuracy. By pooling (mixing of all “good and bad” data) and cumulatively analysing all pressure pairs, the coefficients of correlation, mean prediction and percentage errors (2–9 mm Hg and 2–13 %, respectively) of oscillometry and tonometry indicated almost excellent agreement with invasive radial artery pressure. These data alone, however, may not reveal a bidirectional spread and real discrepancy of noninvasive vs invasive values. One and 2 SD of the errors, and 95 % confidence intervals revealed some disagreement for all patients (table 2, figs 1, 2). Using the absolute error (precision), a rigorous and overall indicator of accuracy, oscillometry showed a narrow SD range of 1–13 mm Hg for systolic, diastolic and mean arterial pressures. Aware of inherent limitations of a CNAP method, the cumulative tonometric precision with an SD range of 6–20 mm Hg could eventually indicate that radial artery tonometry is a clinically useful technique during anaesthesia and surgery. Standards for non-invasive arterial pressure measuring devices and specific requirements during anaesthesia are inadequately defined. They are extrapolated from studies in healthy volunteers or non-surgical patients under steady-state conditions. The American National Standard of the Association for the Advancement of Medical Instrumentation recommended that the maximal bias of non-invasive arterial pressure, obtained from at least 85 patients, should not exceed 5 (8) mm Hg from a non-invasive reference method [30]. The British Hypertension Society expanded the procedure for evaluation of arterial pressure measuring devices, and considered the 5 (8)-mm Hg criterion too liberal. A grading system was proposed according to the percentage of readings 5, 10 and 15 mm Hg from a noninvasive reference method [31]. Both standards and procedures are not applicable directly to anaesthesia and the intraoperative environment, and both are too. British Journal of Anaesthesia rigorous (albeit theoretically desirable) for evaluation of a non-invasive (vs invasive) method of arterial pressure measurements. In this study only the mean pressure value measured oscillometrically (table 2) fulfilled the American standard for minimal bias (SD) [30]. Adopting the British criteria [31], even oscillometric mean arterial pressure (with 50 % of pressure differences within 5 mm Hg and 84 % within 10 mm Hg) would be only partially acceptable. Another approach to the evaluation of noninvasive techniques was considered necessary. To provide more insight into the individual performance of oscillometry and tonometry, the data were pooled and analysed separately for each patient (instead of pooling the data of all patients according to the amount of error). It is a question of judgment if the good or acceptable criteria (table 3) are individually good or acceptable and if systolic, diastolic and mean arterial pressures are of equal importance (as presumed for this analysis) for monitoring high-risk surgical patients. Oscillometry showed a high level of individual accuracy and intermittent monitoring was found to be good or acceptable in all 22 patients. With CNAP monitoring a relevant number of individual pressures (26 %) failed to fulfil our liberal definition of “acceptable” (bias 10 (12) mm Hg, precision 15 mm Hg). The criteria appeared to be too rigorous for systolic pressure (inversely, tonometry particularly failed to identify the peak pressure) but diastolic and mean arterial pressures were also found in the poor category in some patients (table 3). The CNAP method was assessed as a failure in four (18 %) of 22 patients. Adapting the same individual bias criteria to the results of Siegel, Brock-Utne and Brodsky [5], radial artery tonometry was good in six, acceptable in eight and poor in two of their 16 patients, similar to the findings of this study. Most probably, it is not the amount of error per se but the unknown direction of pressure changes which makes tonometric monitoring unreliable (e.g. 2 SD of <27.3 for systolic CNAP indicates a range of 95 % confidence intervals or clinical uncertainty of more than 54 mm Hg) and any subsequent decision for intervention difficult and potentially harmful. Obviously, CNAP monitoring should be interrupted more frequently by oscillometric calibration. It is not clear why tonometric CNAP specifically failed in four of our 22 patients, indicating the general unpredictability of the method during anaesthesia and surgery. The inter-arm pressure difference, obtained before induction of anaesthesia, showed no correlation with oscillometric and tonometric errors obtained later during the procedure. Nevertheless, it should not be dismissed as irrelevant for dynamic comparison over a longer period. The oscillometric and tonometric errors correlated well on average (fig. 3), but the confidence intervals revealed a considerable degree of disagreement between the two methods. It might be assumed that the combination of factors, such as inter-arm pressure difference and a bidirectional relationship of the oscillometric to invasive, oscillometric to tonometric and tonometric to invasive arterial pressure values, individually influenced the accuracy of each non-invasive method..
(7) Radial artery tonometry in high-risk patients. Acknowledgements We thank the Carba-Rüegge Medical Company, Zürich, Switzerland, for providing the Colin instrument.. References 1. Pressman GL, Newgard PM. A transducer for the continuous external measurement of arterial blood pressure. IEEE Transactions on Biomedical Electronics 1963; 10: 73–81. 2. Drzewiecki GM, Melbin J, Noordergraaf A. Arterial tonometry: review and analysis. Journal of Biomechanics 1983; 16: 141–152. 3. Kemmotsu O, Ueda M, Otsuka H, Yamamura T, Winter DC, Eckerle JS. Arterial tonometry for noninvasive, continuous blood pressure monitoring during anesthesia. Anesthesiology 1991; 75: 333–340. 4. Kemmotsu O, Ueda M, Otsuka H, Yamamura T, Okamura A, Ishikawa T, Winter DC, Eckerle JS. Blood pressure measurement by arterial tonometry in controlled hypotension. Anesthesia and Analgesia 1991; 73: 54–58. 5. Siegel LC, Brock-Utne JG, Brodsky JB. Comparison of arterial tonometry with radial artery catheter measurements of blood pressure in anesthetized patients. Anesthesiology 1994; 81: 578–584. 6. Kemmotsu O, Ohno M, Tajita K, Sugimoto H, Otsuka H, Morimoto Y, Mayumi T. Noninvasive, continuous blood pressure measurement by arterial tonometry during anesthesia in children. Anesthesiology 1994; 81: 1162–1168. 7. Weiss BM, Hess OM, von Segesser LK, Bernard E, Hegi T, Pasch T. Transducer with a resonance overshoot eliminator: unpredictable accuracy of a damped systolic arterial pressure. British Journal of Anaesthesia 1995; 74 (Suppl. 1): A27. 8. Sheiner LB, Beal SL. Some suggestions for measuring predictive performance. Journal of Pharmacokinetics and Biopharmaceutics 1981; 9: 503–512. 9. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 307–310. 10. Brunner JMR, Krenis LJ, Kunsman JM, Sherman AP. Comparison of direct and indirect methods of measuring arterial blood pressure (parts I–III). Medical Instrumentation 1981; 15: 11–21, 97–101, 182–188. 11. O’Rourke MF, Avolio A, Kelly RP, Karamanoglu M. Difference between central and upper limb pressure wave forms in man. In: O’Rourke MF, Safar ME, Dzau VJ, eds. Arterial Vasodilation. Mechanisms and Therapy. London: Edward Arnold, 1993; 117–133. 12. Borrow KM, Newburger JW. Noninvasive estimation of central aortic pressure using the oscillometric method for analyzing systemic artery pulsatile blood flow: comparative study of indirect systolic, diastolic, and mean brachial artery pressure with simultaneous direct ascending aortic pressure measurements. American Heart Journal 1982; 103: 879–886. 13. Ramsey M. Blood pressure monitoring: automated oscillometric devices. Journal of Clinical Monitoring 1991; 7: 56–67. 14. Loubser PG. Comparison of intra-arterial and automated oscillometric blood pressure measurement methods in postoperative hypertensive patients. Medical Instrumentation 1986; 20: 255–259. 15. Gersh BJ. Measurement of intravascular pressures. In: PrysRoberts C, ed. The Circulation in Anaesthesia. Applied Physiology and Pharmacology. Oxford: Blackwell Scientific Publications, 1980; 511–529.. 411 16. Gardner RM. Direct blood pressure measurements— dynamic response requirements. Anesthesiology 1981; 54: 227–236. 17. Hipkins SF, Rutten AJ, Runciman WB. Experimental analysis of catheter–manometer systems in vitro and in vivo. Anesthesiology 1989; 71: 893–906. 18. Frank SM, Norris EJ, Christopherson R, Bettie C. Rightand left-arm blood pressure discrepancies in vascular surgery patients. Anesthesiology 1991; 75: 457–463. 19. Coriat P, Vrillon M, Perel A, Baron JF, Le Bret F, Saada M, Viars P. A comparison of systolic blood pressure variations and echocardiographic estimates of end-diastolic left ventricular size in patients after aortic surgery. Anesthesia and Analgesia 1994; 78: 46–53. 20. Rooke GA, Schwid HA, Shapira Y. The effect of graded hemorrhage and intravascular volume replacement on systolic pressure variation in humans during mechanical and spontaneous ventilation. Anesthesia and Analgesia 1995; 80: 925–932. 21. Urzua J, Sessler DI, Meneses G, Sacco CM, Canessa R, Lema G. Thermoregulatory vasoconstriction increases the differences between femoral and radial artery pressures. Journal of Clinical Monitoring 1994; 10: 229–236. 22. Hynson JM, Sessler DI, Moayeri A, Katz JA. Thermoregulatory and anesthetic-induced alterations in the differences among femoral, radial, and oscillometric blood pressures. Anesthesiology 1994; 81: 1411–1421. 23. Ebstein RH, Bartkowski RR, Huffnagle S. Continuous noninvasive finger blood pressure during controlled hypotension. A comparison with intraarterial pressure. Anesthesiology 1991 75: 796–803. 24. Stokes DN, Clutton-Brock T, Patil C, Thompson JM, Hutton P. Comparison of invasive and non-invasive measurement of continuous arterial pressure using the Finapres. British Journal of Anaesthesia 1991; 67: 26–35. 25. Pace NL, East TD. Simultaneous comparison of intraarterial, oscillometric, and Finapres monitoring during anesthesia. Anesthesia and Analgesia 1991; 73: 213–220. 26. Schiller Z, Pasch T. Servo-Plethysmomanometrie zum kontinuierlichen nichtinvasiven Blutdruck-monitoring. Anaesthesist 1991; 40: 105–109. 27. Wilkes MP, Bennett A, Hall P, Lewis M, Clutton-Brock TH. Comparison of invasive and non-invasive measurement of continuous arterial pressure using the Finapres in patients undergoing spinal anaesthesia for lower segment Caesarean section. British Journal of Anaesthesia 1994; 73: 738–743. 28. Young CC, Mark JB, White W, DeBree A, Vender LS, Fleming A. Clinical evaluation of continuous noninvasive blood pressure monitoring: accuracy and tracking capabilities. Journal of Clinical Monitoring 1995; 11: 245–252. 29. Weiss BM, Spahn DR, Keller E, Seifert B, Pasch T. Continuous non-invasive blood pressure monitoring by brachial artery displacement method in high-risk surgical patients. European Journal of Anaesthesiology 1995; 12: 555–563. 30. Association for the Advancement of Medical Instrumentation. American National Standards for Electronic or Automated Sphygmomanometers. Arlington, VA: Association for the Advancement of Medical Instrumentation, 1987, updated 1992. 31. O’Brien E, Petrie J, Littler W, De Swiet M, Padfield PL, O’Malley K, Jamieson M, Altman D, Bland M, Atkins N. The British Hypertension Society protocol for the evaluation of automated and semi-automated blood pressure measuring devices with special references to ambulatory systems. Journal of Hypertension 1990; 8: 607–619..
(8)
Figure
Documents relatifs
Selection of Municipalities is done in consultation with the relevant partners (UNHCR, Serbian Commissioner for Refugees and Ministry of Social Affairs).. Selection of Beneficiaries
En cas de difficultés à respecter les règles trop souvent, nous serons amenés à nous rencontrer afin d’aider votre enfant à adopter une position d’élève nécéssaire pour le
With regard to the other progestogens, the ESTHER (EStrogen and THromboEmbolism Risk) Study, a French case/control study of VTE among postmenopausal women, was
The precipitons move according to hydrodynamic equations, and modify topography by erosion and deposition laws dependent on both hydraulic conditions (i.e., shear stress,
He rejected as being dangerous to theology the attempts of theologians such as James McCosh, the president of Princeton University, to reconcile theory and
This note examines in some detail the design snow loads on simple arches and curved roofs recommended in the 1977 Commentary on Snow Loads of the National Building Code of
Cette étude expérimentale par une méthode originale, a permis de cartographier sur les valeurs de module de conservation en fonction de la température et de la contrainte. Après
Finally, as specified in the CHMP guideline on strategies to identify and mitigate risks for FIH studies and early clinical trials with investigational medicinal products ( 22 ),
