Comparison of Automated and Mercury Column Blood Pressure Measurements in Health Care Settings.

OBJECTIVES: Use of automated (electronic) blood pressure measurement devices to obtain clinic blood pressure measurements is becoming increasingly widespread in health care settings; their comparability with manual mercury sphygmomanometer readings is uncertain. Current guidelines for screening, diagnosis, and treatment of hypertension in clinical practice are based on clinical trials and epidemiological evidence derived from readings taken with auscultatory devices (usually mercury sphygmomanometers). This study was carried out to assess whether use of automated oscillometric devices in the clinical setting led to differences in classification of blood pressure levels at the 140 mm Hg systolic and 90 mm Hg diastolic thresholds compared to readings obtained with a standard mercury sphygmomanometer. DESIGN: We compared the blood pressure readings obtained with three automated devices that are widely available in the U.S. (one Dinamap Plus Model 8710 and two Dinamap Model 1846SX) and the readings taken by a single trained research assistant with a manual mercury column device in the emergency department and the outpatient medicine clinic of a large urban teaching hospital. The devices tested were those in normal use in the setting. The order in which the readings were taken was varied randomly. The sensitivity and specificity of the Dinamap readings compared to the mercury column device as the gold standard was calculated. RESULTS: The mean diastolic blood pressure in the three groups of patients studied was 7.3 mm Hg, 2.4 mm Hg, and 3.4 mm Hg lower with the Dinamap devices than the mercury column device (p is less than 0.001 for all comparisons). The mean systolic blood pressure readings were 1.0 mm Hg (p equals 0.06), 6.7 mm Hg (p is less than 0.001), and 4.2 mm Hg (p is less than 0.001) higher with the Dinamap device than the mercury column device. The difference between Dinamap and mercury column systolic blood pressure readings tended to increase at pressures greater than 140 mm Hg, whereas the diastolic blood pressure differences remained uniform throughout the blood pressure range. The sensitivity of the Dinamap readings compared to mercury column readings for classifying individuals as hypertensive was 73% for persons with elevated systolic blood pressure only, 51% for persons with both systolic and diastolic blood pressure elevation, and 10% for person with diastolic blood pressure elevation only. CONCLUSIONS: The Dinamap devices tested appeared to yield systematically biased blood pressure readings, which could alter the assessment of hypertension prevalence and control in clinical populations. Quality assurance and outcome researchers should attempt to document the type of device used to obtain blood pressure measurements noted in medical records, and be cognizant that small systematic errors in measurement could affect a setting's hypertension control performance. (c)2000 by Le Jacq Communications, Inc.