Reproducibility of non-invasive and intra-arterial blood pressure monitoring: implications for studies on antihypertensive treatment

Abstract

Ambulatory blood pressure has been shown to be more reproducible than office blood pressure and thus to be more suited for studying the efficacy of antihypertensive drugs. In 34 untreated essential hypertensive subjects, we measured office and 24-h non-invasive or intra-arterial blood pressure twice over a 4-week interval; 24-h intra-arterial blood pressure was obtained by the Oxford method whereas 24-h non-invasive blood pressure was obtained by the automatic SpaceLabs 5300 device, with a 10min (daytime) or 20min (night-time) interval between measurements. The standard deviation of the mean difference (s.d.d.) between blood pressures obtained in each recording was taken as the reciprocal of blood pressure reproducibility. The s.d.d. was highest for office blood pressure and for single blood pressure readings taken from 24-h non-invasive recordings. The s.d.d. fell when the two 24-h average non-invasive blood pressures were considered. The fall was progressively greater as the number of ambulatory readings on which the average was calculated increased from two to 24, no further fall being observed when more than 24 values were considered. The maximal reduction in s.d.d. was 59% (systolic) and 42% (diastolic) as compared with the office s.d.d. The two 24-h mean values obtained by the intra-arterial approach were slightly more closely correlated than those obtained non-invasively. However, at comparable sampling rates, the s.d.d. was not substantially lower with 24-h intra-arterial blood pressure and including in the calculation the average of the thousand readings provided by this approach did not cause any further improvement. Thus, 24-h average blood pressure is more reproducible than office blood pressure. The increased reproducibility is a function of the number of ambulatory blood pressure readings, but a maximal advantage can be obtained at a relatively low sampling rate. This means that in therapeutic trials there may be no need to assess blood pressure too frequently or to consider the huge number of values provided by intra-arterial monitoring