[How reliable is conventional blood pressure registration? Comparison with a semi-automatic device]

The aim of the present study was to determine observer error in measuring blood pressure by the conventional auscultatory method. Casual blood pressure was measured in two age-matched groups of normo- and hypertensive patients with a conventional mercury sphygmomanometer (n = 181) or a semiautomatic device (n = 176) by ten doctors of the university hospital. Although there were no significant differences in mean systolic or diastolic blood pressure values between the groups, the conventionally determined values showed a distinctively different pattern of distribution in comparison to the semiautomatically taken readings. With conventional readings a highly significant preference for terminal digit "0" (44%) and fewer systolic and diastolic values in the lower blood pressure range were observed. Furthermore there was a higher frequency of conventional readings ending in "8" than in "2". Terminal digit preference in the whole group was mainly due to the doctors, who did not measure blood pressure to the nearest 2 mmHg mark. Our results thus stress the importance of better and regular training in the correct technique of measuring blood pressure in order to reduce sources of observer bias.