Comparison of performance of various sphygmomanometers with intra-arterial blood-pressure readings.

Seven types of sphygmomanometer were used in random order on each of nine hypertensive patients and the readings compared with simultaneous intra-arterial blood-pressure recordings. All the devices gave significantly different values for systolic pressure, and only two measured diastolic pressure without significant error. Systolic pressure was consistently underestimated (range 31-7 mm Hg), and all but one instrument overestimated diastolic pressure (range 10-2 mm Hg). The variability of readings was least with the standard mercury sphygmomanometer and the random-zero machine, while with some of the more automated devices single readings were in error up to -68/33 mm Hg. The strong correlations found between intra-arterial and cuff systolic pressures with all devices tested and significant correlations for diastolic pressure with all but one device indicate that, with one possible exception, the sphygmomanometers would give accurate results where a change in blood pressure was the main concern.