Osler's maneuver: Absence of usefulness for the detection of pseudohypertension in an elderly population

Abstract

To investigate the prevalence and clinical relevance of Osler's maneuver for detecting pseudohypertension in the elderly. Osler's maneuver was performed by one investigator in 205 consecutive inpatients of a French geriatric hospital (40 men and 165 women; mean age 84.2 +/- 6.2 years). In 12 Osler-positive and 12 Osler-negative patients matched for age, sex, and presence of hypertension, the blood pressure values measured directly at the radial artery were compared to those measured indirectly with a standard mercury sphygmomanometer. Aortic and upper limb pulse wave velocities (PWV) were also measured in these 24 patients. Twenty-three of 205 patients (11%) were Osler-positive. Age, sex ratio, and prevalence of hypertension or other cardiovascular diseases did not differ significantly in Osler-positive and Osler-negative patients. Systolic blood pressure (SBP), measured by standard mercury sphygmomanometer, was significantly greater in Osler-positive than Osler-negative patients (157 +/- 37 versus 132 +/- 28 mm Hg; P < 0.01). Diastolic blood pressure (DBP) did not differ significantly (78 +/- 18 versus 74 +/- 14 mm Hg). Interobserver agreement concerning Osler's sign, studied in 40 patients, was good (kappa = 0.72). In 12 Osler-positive and 12 Osler-negative patients, the mean differences between SBP obtained by cuff-manual indirect blood pressure and direct measurements were -3.71 +/- 22.85 mm Hg and -8.59 +/- 14.40 mm Hg (P = NS). For DBP, these differences were 18.40 +/- 15.72 and 12.01 +/- 5.80 mm Hg (P = NS). The differences between the indirect and direct blood pressure measurements were significantly correlated to upper limb PWV, but not aortic PWV, for both SBP and DBP. Pseudohypertension, defined as the indirect measurement overestimation of SBP or DBP by 10 mm Hg or more, was found in 15 of the 24 patients (63%). In these patients, upper limb PWV was significantly greater than in those with no pseudohypertension (7.0 +/- 2.2 versus 5.4 +/- 1.3 m/s; P < 0.05). The Osler-positive maneuver is frequently found in elderly hospital inpatients, but its ability to detect pseudohypertension in clinical practice is poor. Measurement of upper limb PWV might be a more appropriate way of screening for this condition.