We all are trained to measure our patients' vital signs ourselves. The process of holding your patient's arm close to your side with your elbow, placing the cuff, pumping it up, and deflating it while listening to the brachial artery with your stethoscope adds an intimacy to the physical exam and a way of slowing down an otherwise hectic encounter.
However, most vital signs are now obtained using automated techniques, thanks to the necessity of repetitive measurements and caregivers' limited time. The nurse or aide rolls the noninvasive machine in, places the cuff on the patient's arm, pushes a button, records the numbers, and rolls off to the next patient. The doctor or allied health provider looks at the numbers and decides if any therapeutic action is required. We put a lot of stock in these numbers. Just how accurate are they?
Most noninvasive blood pressure monitors use the oscillometric technique. The cuff is placed on the patient's arm, and the cuff bladder is inflated with air until the external pressure exceeds the intra-arterial systolic pressure and arterial flow past the cuff ceases. The cuff bladder pressure is slowly released. A pressure sensor inside the cuff detects arterial pulsations as oscillations. As the cuff pressure declines, the oscillations increase in amplitude to a maximum, which represents the mean arterial pressure (MAP). The oscillation amplitude then declines until minimal. The only pressure actually measured is the pressure at the point of maximal amplitude, which corresponds to the MAP.
The computer uses an algorithm or parameters designated by the manufacturers to produce the systolic and diastolic blood pressure readings. For example, the systolic and diastolic pressures may be calculated as the pressures at which the amplitudes are a fixed percentage of the maximum oscillation. Each manufacturer has its own method, meaning there may be considerable variation between machines.
Many factors can affect the accuracy of the oscillometric technique (see sidebar). Anything altering the oscillation amplitude or regularity will produce erroneous results. Increased arterial wall stiffness due to arteriosclerosis, usually seen with increasing age, is associated with overestimated systolic (mean, 10.9 mm Hg) and diastolic (mean, 4.8 mm Hg) pressures (Hypertension. 2000;36;484-488). O’Brien and Atkins reviewed 10 studies of the Dinamap 8100 and found that only 66% of systolic measurements came within 5 mm Hg of the mercury sphygmomanometer, and fewer than 50% of diastolic measurements came within 5 mm Hg of the mercury standard (Lancet. 1997;349:1026). Severe peripheral vascular disease makes oscillometric measurements unreliable in the lower extremities (Clin Physiol. 2001;21:155-163). A study in lean and obese patients found inaccuracies regardless of body weight or arm circumference (Angiology. 2006;57:41-45). Finally, increasing pulse pressure and small arm circumference were associated with unreliable oscillometric blood pressure measurements in 5,000 measurements of 755 patients over time (J Hum Hypertens. Epub 26 March 2009).
To summarize, automated noninvasive blood pressure and pulse measurements are here to stay. It is important to recognize the potential for inaccuracies, and attempt to confirm abnormal or unexpected measurements before plunging ahead with treatment.