Section 3: Anesthesia Management
Part B: Monitoring
Chapter 30: Cardiovascular Monitoring

Indirect Measurement of Arterial Blood Pressure

Manual Intermittent Techniques

Most indirect methods of blood pressure measurement rely on a Riva-Rocci sphygmomanometer. As described by Riva-Rocci in 1896, 25  this apparatus included an armencircling inflatable elastic cuff, a rubber bulb to inflate the cuff, and a mercury manometer to measure cuff pressure. 26  Riva-Rocci described the measurement of systolic arterial blood pressure by determining the pressure at which the palpated radial arterial pulse disappeared as the cuff was inflated. The scientific rigor and attention to detail of Riva-Rocci‘s work are chronicled in a translation celebrating his original publications. 26 

A variation of the Riva-Rocci method commonly employed today is generally termed the return-to-flow technique. Whereas the Riva-Rocci technique recorded the pressure during cuff inflation at which the pulse completely disappeared, the return-to-flow method records the pressure during cuff deflation at which the pulse reappears. Using this technique, systolic blood pressure can be estimated without a stethoscope using only a cuff and manometer. When the patient has a finger pulse oximeter or indwelling arterial catheter in the ipsilateral arm, the return to flow can be determined visually by the reappearance of the plethysmographic or arterial pressure waveforms.

Although return-to-flow methods provide a simple, rapid means to estimate systolic blood pressure, they do not allow measurement of diastolic blood pressure. Undoubtedly, the most widely used intermittent manual method for blood pressure determination is the auscultation of sounds originally described by Korotkoff in 1905. 27, 28  Using a Riva-Rocci sphygmomanometer and cuff, Korotkoff applied a stethoscope to the artery directly below the cuff to auscultate the sounds generated as the cuff was slowly deflated. These sounds are a complex series of audible frequencies produced by turbulent flow, instability of the arterial wall, and shock wave formation created as external occluding pressure on a major artery is reduced. 24  The pressure at which the first Korotkoff sound is auscultated is generally accepted as the systolic pressure (phase I). The sound character progressively changes (phases II and III), becomes muffled (phase IV), and finally absent (phase V). Diastolic pressure is recorded at phase IV or V. However, phase V may never occur in certain pathophysiologic states, such as aortic regurgitation. 29 

The auscultatory method for blood pressure determination is limited by excessively long or loose stethoscope tubing, which impairs sound transmission, or by poor hearing sensitivity of the observer. Aneroid manometers are subject to calibration errors and should be checked periodically. A more basic shortcoming of auscultation is the reliance on blood flow to generate Korotkoff sounds. Pathologic or iatrogenic causes for decreased peripheral blood flow, such as cardiogenic shock or high-dose vasopressor infusion, can attenuate or obliterate sound generation and result in significant underestimation of blood pressure. 30  In contrast, low compliance of the tissues underlying the cuff, as encountered in a shivering patient, will require an excessively high cuff-occluding pressure and produce “pseudohypertension.” 24  Patients with severe calcific arteriosclerosis may have relatively noncompressible arteries, another circumstance wherein cuff blood pressures will yield spuriously elevated results compared with true intra-arterial blood pressure. 31 

Other common sources of error during intermittent manual blood pressure measurement include selection of an inappropriate cuff size or excessively rapid cuff deflation. Blood pressure cuff width should be 20 percent greater than arm diameter, and the cuff should be applied snugly after any residual air has been squeezed out. The pneumatic bladder inside the cuff should span at least half the circumference of the arm and be centered over the artery. Although too large a cuff generally will work well and produce little error, cuffs that are too narrow yield erroneously elevated values for blood pressure (Fig. 30–2). 32, 33  Cuff deflation rate is another important variable that influences accurate blood pressure measurement, especially when deflation is performed manually. The decrease in cuff pressure should proceed slowly enough for the Korotkoff sounds to be auscultated and assigned properly to the current pressure in the cuff. Failure to identify the initial Korotkoff sounds will result in a falsely low measure of blood pressure. A deflation rate of 3 mm Hg/s limits this source of error, and coupling deflation rate to heart rate—2 mm Hg/beat—has been found to improve accuracy further. 34 

Click thumbnail to see full size image
FIGURE 30–2 Effect of cuff size on manual blood pressure measurement. An inappropriately small blood pressure cuff yields erroneously high values for blood pressure because the pressure within the cuff is incompletely transmitted to the underlying artery. See text for greater detail.

Automated Intermittent Techniques

Many limitations of manual intermittent blood pressure measurement have been overcome by automated noninvasive blood pressure (NIBP) devices, which are now used widely in medical care. By applying a single algorithm or method of data interpretation, NIBP devices provide consistent, reliable values for systolic, diastolic, and mean arterial pressure (MAP). In addition, automated NIBP devices provide alarm systems to draw attention to extreme blood pressure values and have the capacity to transfer data to an automated trending device or information system. However, the greatest advantage of automated NIBP devices compared with manual methods of blood pressure measurement is that they free the operator to perform other vital clinical duties while simultaneously ensuring frequent, repetitive blood pressure measurements.

Most automated NIBP devices are based on the technique termed oscillometry, a technique first described by von Recklinghausen in 1931. 35  In this method, variations in cuff pressure resulting from arterial pulsations during cuff deflation are sensed by the monitor and used to determine arterial blood pressure values. Peak amplitude of arterial pulsations corresponds closely to true MAP. 36, 37  (Interestingly, the reason many clinicians still refer to an automated NIBP device as a “Dinamap” is that the original clinically available monitors provided only this single, easily measured pressure value—MAP.) Values for systolic and diastolic pressure are derived using proprietary formulas that examine the rate of change of the pressure pulsations (Fig. 30–3). Systolic pressure is generally chosen as the pressure at which pulsations are increasing and are at 25 to 50 percent of maximum. Diastolic pressure is more difficult to determine but is commonly placed at the point where the pulse amplitude has declined by 80 percent. 24 

Click thumbnail to see full size image
FIGURE 30–3 Comparison of blood pressure measurements using Korotkoff sounds and oscillometry. Oscillometric systolic blood pressure is recorded at the point where cuff pressure oscillations begin to increase, mean pressure corresponds to the point of maximal oscillations, and diastolic pressure is measured where the oscillations become attenuated. Note the correspondence between these measurements and the Korotkoff sounds that determine auscultatory systolic and diastolic pressures. (From Geddes LA: Cardiovascular Devices and Their Applications. New York; John Wiley, 1984: Fig. 34–2. Reprinted by permission of John Wiley & Sons, Inc.)

Although oscillometry is used primarily in automated NIBP measurement, the same principles may be applied to determine blood pressure manually using a standard cuff and aneroid manometer. If the cuff is deflated slowly until the needle on the aneroid gauge begins to flicker or oscillate, this pressure value will provide a close estimate for systolic blood pressure. The primary advantage of this technique is that it can be performed quickly, using only one hand on the inflation bulb and pressure relief valve, thereby leaving the anesthesiologist‘s other hand free to maintain the patient‘s airway.

In clinical practice, oscillometric automated NIBP measurement has focused largely on pressures measured from the upper arm. Sometimes the patient‘s surgical procedure or medical condition requires that the cuff be applied to the calf, ankle, or thigh. These are all acceptable alternatives, keeping in mind that an appropriate size cuff must be employed. Because the oscillometric NIBP technique provides accurate pressure measurements in pediatric patients, 38, 39, 40  several investigators have proposed using neonatal sized cuffs placed around a finger or thumb of an adult patient. 41, 42, 43  Although these alternatives may be appropriate in certain circumstances, their overall accuracy has not been widely validated and may not conform to accepted industry standards. 44 

Many other techniques have been described for automated NIBP measurement, but none have supplanted the standard oscillometric technique. One device uses the Doppler principle to determine blood flow distal to the cuff 45  and another senses motion of the arterial wall. 46  Although these have been found to record blood pressure with acceptable accuracy, they require extra efforts to place and stabilize additional sensing transducers. Supraorbital artery oscillometry has been studied as an alternative site for blood pressure measurement but appears inaccurate compared with direct radial artery pressure values, perhaps owing to the peripheral location and vasoreactive sensitivity of the supraorbital artery. 47  Photo-oscillometry to detect brachial artery movement has also been described, but awaits broader validation in clinical practice. 48 

Under controlled clinical conditions, numerous investigators have demonstrated that automated NIBP measurements closely approximate directly measured arterial pressure. 22, 32, 37, 49  However, other studies underscore the fact that marked disagreement occurs when direct and indirect pressure measurements are compared, 50  particularly when radial artery pressure is used as the direct measurement standard 23, 51  or when techniques are compared under changing clinical conditions. When direct brachial artery pressures have been compared with various indirect methods, (including manual auscultation, automated oscillometry, aneroid manometer visual onset of needle oscillations, and return to flow), the relation between the indirect and direct pressures varied between patients, within patients over time, and with changing hemodynamic conditions. 52 

As noted, some authors have emphasized the lack of exact agreement between different measurements of blood pressure. 23, 24, 48, 53  Standards for performance of automated NIBP devices have been advanced by organizations such as the American Association for the Advancement of Medical Instrumentation (AAMI) and the British Hypertension Society. AAMI standards require a monitor to record blood pressure to within 5 ± 8 mm Hg (mean ± standard deviation) prediction error compared with the reference method. 48  However, clinical performance of an NIBP monitor should be evaluated by other criteria as well. These include the number of outlier values, duration of discrepancies, magnitude of individual errors, and performance under variable clinical conditions. 48 

Although automated blood pressure measurement techniques are considered noninvasive and relatively safe, complications have been reported. These include pain, petechiae and ecchymoses, limb edema, venous stasis and thrombophlebitis, peripheral neuropathy, and even compartment syndromes. 48, 54, 55, 56  These morbid events occur more often following prolonged periods of excessively frequent cuff inflation/deflation cycling, resulting in local trauma or impaired distal limb perfusion. Other factors that may contribute include cuff misplacement across a joint or repeated attempts to determine blood pressure in the presence of an artifact-producing condition, such as involuntary muscle tremors. 57  Caution should be exercised when using these monitors in patients with depressed consciousness, preexisting peripheral neuropathies, arterial or venous insufficiency, or irregular cardiac rhythms. 54 

Automated Continuous Techniques

In the past, continuous blood pressure monitoring required direct arterial cannulation. Advances in microprocessor and servomechanical control technology have enabled noninvasive techniques to provide a reasonable representation of the arterial pressure waveform and a nearly continuous assessment of blood pressure. The most widely investigated of these devices measures finger blood pressure using a servoplethysmomanometer, designed and first reported by Penaz in 1973. 58 

In essence, the finger blood pressure device uses an arterial volume-clamp method. It consists of a small cuff secured around the middle phalanx of a finger or the base of the thumb. The inflatable, flexible cuff contains an infrared photoplethysmograph. The cuff and photoplethysmograph are linked through a sophisticated servocontrolled mechanism housed within a small box that is strapped to the wrist. The plethysmograph continually measures the size (i.e., diameter) of the digital arteries using transillumination. To begin monitoring, a “locking” calibration procedure is performed by varying cuff pressure to establish the vessel size at which oscillometric pressure variation is maximal. As noted, this corresponds to MAP. An electromechanical feedback loop is then established, and external pressure applied to the cuff is varied continuously to keep the measured vessel size constant at the set-point. Thus, cuff pressure tracks arterial pressure throughout the cardiac cycle and is displayed on the monitor screen as a continuous waveform. More detailed descriptions of this technology are readily available. 59, 60, 61, 62, 63 

Over the past 15 years, continuous noninvasive finger blood pressure measurement devices have undergone numerous clinical evaluations comparing their performance with direct arterial pressure measurements. 59, 60, 61, 62, 63  Many of these investigations have demonstrated small overall mean differences between finger and intra-arterial pressure meas-urements. However, as noted earlier, small overall mean pressure differences do not necessarily indicate good meas-urement agreement. In clinical practice, the frequency of major measurement errors may be the more relevant and important issue. For example, Smith et al 63  found that finger arterial spasm precluded accurate pressure measurement in 5 percent of patients. Gibbs et al 64  studied 20 patients during anesthesia for major surgery and reported that MAP measured from a finger cuff differed from direct arterial pressure values by more than 10 mm Hg in 30 percent of patients and more than 20 mm Hg in 5 percent of patients. Other investigations in patients undergoing thoracic 65  and vascular 66  surgery similarly have shown small average pressure differences overall but large discrepancies in a significant number of individual patients.

In addition to the questionable accuracy of finger blood pressure measurement, other factors undoubtedly have limited more widespread application of this technology. Unlike direct invasive arterial blood pressure monitoring, indirect noninvasive pressure measurement does not provide access for blood sampling, which is a frequent requirement in the setting of continuous pressure monitoring. By definition, finger blood pressure monitoring records a distal arterial pressure, which tends to be lower than brachial arterial pressure in elderly patients with atherosclerosis and to be higher than brachial pressure in young patients, because of peripheral pulse wave amplification. Using sophisticated electronic processing techniques, Bos et al 67  demonstrated that the pulse wave shape and pressure values of the finger blood pressure signal could be corrected to resemble direct brachial artery pressure waveforms within the limits of accuracy described by the AAMI. However, these sophisticated algorithms have not yet been integrated into commercially available monitors. Because the “transducer” recording blood pressure with this method is the finger, the vertical height of the finger becomes an important determinant of the pressure recorded, just as transducer height is important with direct arterial pressure measurement (see later discussion). 63  Finally, the potential for circulatory impairment of the distal finger caused by the constantly inflated cuff has been a cause for concern. Gravenstein et al 68  demonstrated mild hypoxemia in the capillary blood of the fingertip during finger blood pressure monitoring. No adverse outcomes were noted in these study patients or in others in whom finger blood pressure measurement was performed for as long as 7 hours. 63  Despite the apparent safety of these devices for short-term use, it is not surprising that these considerations have limited more widespread clinical application of continuous finger blood pressure monitoring.

Other automatic and continuous techniques have been used to measure blood pressure noninvasively. One such device reconstructs an arterial pressure waveform from arterial wall displacement measurements, following an oscillometric calibration. Unfortunately, changes in arterial compliance affect the clinical performance of this instrument. 69  Because arterial pulse wave velocity depends on arterial blood pressure, another device uses pulse transit time as recorded from dual pulse oximeter probes placed on the ear and finger following oscillometric calibration from the contralateral arm. 70  This device has also performed poorly in clinical settings, with indirectly monitored blood pressures changing in an opposite direction from the direct intraarterial pressure more than 30 percent of the time. 70  A third method used for continuous noninvasive pressure monitoring is arterial tonometry, a version of applanation tonometry. 71, 72, 73  In brief, a superficial artery (usually the radial) is compressed and partially flattened against the underlying bone. This flattened arterial surface serves as a “transducer” for intravascular pressures acting perpendicularly against the vessel wall. An array of piezoelectric crystals positioned on the skin overlying this flattened portion of artery senses arterial pressure changes and translates them into a continuous arterial pressure waveform. The device is calibrated at intervals by cuff oscillometry from the upper arm. Early investigations suggested that the clinical performance of this device was better than other forms of continuous noninvasive pressure monitoring. 71, 72  Unfortunately, more recent clinical studies have identified limitations of this device in pediatric patients 74  and in patients receiving vasodilating drugs. 48  Like other automated, continuous noninvasive blood pressure techniques, arterial tonometry holds promise as a means of rapidly detecting changes in blood pressure. However, in terms of absolute accuracy of blood pressure measurement, intermittent oscillometry appears to be superior to radial artery tonometry. It remains unclear whether any noninvasive technique will significantly reduce the need for direct arterial pressure monitoring or whether these methods will replace automated intermittent oscillometry as the standard noninvasive blood pressure monitoring method in anesthesia and critical care.