Evaluation of an automated blood pressure measuring device intended for general public use

Responding to Chicago newspaper reports, measurements of blood pressure by a publicly available, automated coin-operated device were compared with those of human observers using the standard cuff and auscultatory technique. One machine was examined in the laboratory, and eight others at randomly selected sites. Analysis of readings made on 100 persons in the laboratory and 227 in the field led to the following conclusions: 1) On the average, the machines measured fifth phase diastolic blood pressure at nearly the same level as did human observers; 2) The machines were more variable measuring systolic blood pressure with four differing from the average human reading by 1mm Hg or less, but two differing by 8mm Hg or more; 3) The agreement between machine-human pairs of readings was not as good as between human-human pairs, but the differences in level of agreement--both in determining the actual value and in categorizing the values as normal, borderline, or high--were small and have little practical importance; 4) Linear regression analyses of the relationship between simultaneously determined machine and human readings indicated that the average human-machine difference was the same over the range of pressures tested. Publicly available blood pressure measuring devices should be labeled concerning their purposes, capabilities, and limitations. Rules and regulations governing their use in the City of Chicago are being prepared by this city's Legal Department.