Is the accuracy of blood pressure measuring devices underestimated at increasing blood pressure levels?

Background In validation studies reporting on the accuracy of blood pressure measuring devices (ambulatory and non-ambulatory systems), it is frequently stated that the accuracy of blood pressure devices seems to decrease at increasing blood pressure levels. This has been shown for several ambulatory devices in the past. Whether more recently validated devices are less accurate at increasing blood pressure levels is unknown, however. Objectives We therefore retrospectively searched the literature for studies performed between 1993 and 2003, reporting on the accuracy of blood pressure measuring devices over different blood pressure levels. When needed, additional information from the authors was requested. Methods In total, 30 studies were selected. Of these, the studies reporting on the accuracy of 14 different ambulatory and nine different non-ambulatory devices were useful. For both ambulatory and non-ambulatory devices, accuracy appeared to decrease at increasing blood pressure levels. This was particularly shown for systolic blood pressure. Results We speculate whether this finding is due to the oscillometric method of blood pressure measurement. Another explanation may exist, however. Blood pressure variability increases with higher blood pressure. Further, the British Hypertension Society protocol 1993 uses sequential measurements. This may be the reason that, owing to the increased blood pressure variability, the accuracy of most devices tends to decrease at higher blood pressure levels. Consequently, the accuracy of blood pressure measuring devices may be underestimated at higher blood pressure levels. Conclusion Currently used automated blood pressure measurement devices seem to be less accurate at increasing blood pressure levels. It is important to be aware of this phenomenon when treating hypertensive patients. The reported decrease in accuracy, however, may well be explained by the increasing blood pressure variability at increasing blood pressure and the use of sequential measurements. If this is the case, then the accuracy of these devices is perhaps underestimated.