Patient-specific differences between blood pressure estimated from 24 h ambulatory measurements and serial office self-recordings

The long-term true blood pressure, estimated in terms of the repeatable 24 h ambulatory mean (24 hAmb) pressure, is probably best related to progressive cardiovascular deterioration and is vital for clinical decision-making. Serial self-measurements (SELF) under quiet conditions may reflect this level, but with an annoying uncertainty in individual patients. This uncertainty is characterized by analyzing differences between the SELF readings of seated/supine patients on their own and the reference 24 hAmb values based on 72 recordings. From at-random replicated sessions of 59 subjects, the sources of uncertainty can be separated into three components: a systematic (non-random) variable difference in level from person to person, a mean systematic difference in level between methods, and random variability determining repeatability between sessions. An unstable alerting reaction distorts about six initial self-recordings and increases the random variability. The following 7-12 or 13-24 'steady-state' values show a comparable random variability but reflect the 24 hAmb level only when compared as the average of all patients. The standard deviation for repeatability is about 50-100% higher for SELF than for 24 hAmb. The patient-specific difference between methods contributes more than +/-10 mmHg to the uncertainty interval of the SELF results. These patient-specific differences recur at repeat sessions and thus cannot be reduced by averaging the results of sequential SELF sessions. In contrast, two 24 hAmb results provide a 95% uncertainty interval of +/-5 mm Hg. Thus, the averaged results of multiple SELF sessions can be used to detect major changes in blood pressure but are, as a result of patient-specific differences, too unreliable when the value to be estimated (the true value) is the expected long-term value of 24 hAmb results.