Blood pressure evaluation by noninvasive and traditional methods. Consistencies and discrepancies among photoplethysmomanometry, office sphygmomanometry, and ambulatory monitoring. Effects of blood pressure measurement.

Three hundred eight outpatients referred for hypertension were studied. A continuous beat-to-beat noninvasive recording (Finapres) of blood pressure evaluated the blood pressure increase (9 mm Hg systolic and 4 mm Hg diastolic) induced by office sphygmomanometry. Thereafter, patients underwent a 24-h ambulatory blood pressure monitoring. The evaluation against Finapres showed that office sphygmomanometry overestimates the systolic blood pressure by 3 +/- 36 mm Hg (mean +/- 2 SD) and the diastolic blood pressure by 15 +/- 25 mm Hg (mean +/- 2 SD). Blood pressure monitoring showed similar discrepancies. On the basis of both monitoring (normalcy defined from a population of 550 normotensive subjects) and office sphygmomanometry, patients were considered normotensive, hypertensive (either untreated or under active drug treatment), white coat hypertensive (monitoring below the 95th percentile and sphygmomanometry more than 140/90 mm Hg, either untreated or under active drug treatment), and reverse white coat patients (monitoring over the 95th percentile and sphygmomanometry less than 140/90 mm Hg). Patients showed different levels of alert reaction (the highest in white coat hypertensive and the lowest in reverse white coat hypertensive patients), and a similar increase in blood pressure induced by conventional sphygmomanometry. During initial readings of ambulatory monitoring, blood pressure decreased from the first reading to the third reading. This decrease is related to the increase of blood pressure under sphygmomanometry. Caution should be paid in interpreting results of sphygmomanometry (error level in the single patient as high as +/- 40 mm Hg), and interpreting and averaging results of the first hour of blood pressure monitoring (variably affected by the alert reaction to the clinical environment).