Performance of a new blood pressure monitor in children and young adults: The difficulties in clinical validation

Abstract

The aim of this study was to assess the performance of a new ambulatory blood pressure monitor (Dynapulse® 5000A) in children and young adults. The reference methods, i.e. the mercury-Doppler (Hg-D) method of blood-pressure measurement (recommended for less than 5-year-old children by the British Hypertension Society criteria) and the mercury-auscultation (Hg-A) technique (recommended for 5 years and above)were initially compared utilizing 132 concomitant systolic blood pressure (SBP) readings taken among 44 subjects (3 in each) aged 2-30 years. Hg-D over-read SBP above that of the conventional Hg-A by a mean of 5.25mmHg (95% confidence interval (CI) 4.11, 6.4). This difference in measurement between the two techniques was also significantly (p = 0.01) related to age. [Difference in mmHg = 7.83 (95% CI 5.53,10.12) - 0.30 (95% CI -0.54, -0.07) age in years.] The Dynapulse® 5000A was therefore validated only against Hg-D throughout the age range and was undertaken utilizing 468 concomitant resting SBP measurements taken among 59 subjects (supine or sitting) using a 'T' system. The Dynapulse was found to give higher SBP readings (average 13.5 mmHg) than Hg-D sphygmomanometry. Although the differences were not dependent on the age of the child, the pulse rate or the body mass index, there was a significant drift in the differences with time (-0.05 mmHg/day). Repeatability of each of the measurement methods calculated utilizing two or three consecutive recordings taken at 1-min intervals showed that variability between repeated measurements made with the Dynapulse® (14.8%) was greater than for Hg-D. Repeatability coefficient of the Hg-D (11.2%) was also less than ideal for a reference method. We conclude that Dynapulse 5000A over-reads SBP above that detected by Hg-D, but cannot be fairly graded by British Hypertension Society criteria due to the inconsistency and poor repeatability of the recommended reference methods.