Accuracy of a piezoelectric crystal microphone interfaced through an audio amplifier board for use in an automated blood pressure device.

Two certified blood pressure technicians and a computerized blood pressure monitor from Computerized Screening, Inc., model 2000 (CSI 2000) obtained 900 blood pressure (BP) readings from 100 subjects aged 15 to 85 years (mean = 44.1 +/- 1.8 SEM). Measurements were obtained simultaneously on the same limb by the observers and the CSI 2000 three times per subject within a 30-minute period). Pressures obtained from the CSI 2000, which uses a piezoelectric crystal microphone interfaced through a newly designed audio amplifier board, were compared with the results of indirect cuff mercury sphygmomanometry. Correlations between the CSI 2000 readings and simultaneous readings obtained by two certified BP technicians for systolic and diastolic BP were r = 0.96 (p less than 0.001) and r = 0.91 (p less than 0.001), respectively. For systolic BP readings, the estimating equation (least-squares) based on a linear relationship between measurements from the test and reference methods was: CSI 2000 systolic BP = -3.454 + [1.015 x (observers' systolic BP)]. For diastolic BP readings, the estimating equation (least squares) was: CSI 2000 = 5.997 + [0.906 x (observers' diastolic BP)]. Close agreement was found between the readings obtained by the CSI 2000 and the average of those obtained by the two certified BP technicians.