Institutional members access full text with Ovid®

Share this article on:

Discordant prevalence of hypertension using two different automated blood pressure measurement devices: a population-based study in Dar es Salaam (Tanzania)

Chiolero, Arnauda b; Gervasoni, Jean-Pierrea; Rwebogora, Annec; Mkamba, Mashomboc; Waeber, Bernardb; Paccaud, Freda; Burnier, Michelb; Bovet, Pascala

Clinical Methods and Pathophysiology

Objective The estimation of blood pressure is dependent on the accuracy of the measurement devices. We compared blood pressure readings obtained with an automated oscillometric arm-cuff device and with an automated oscillometric wrist-cuff device and then assessed the prevalence of defined blood pressure categories.

Methods Within a population-based survey in Dar es Salaam (Tanzania), we selected all participants with a blood pressure ≥160/95 mmHg (n=653) and a random sample of participants with blood pressure <160/95 mmHg (n=662), based on the first blood pressure reading. Blood pressure was reassessed 2 years later for 464 and 410 of the participants, respectively. In these 874 subjects, we compared the prevalence of blood pressure categories as estimated with each device.

Results Overall, the wrist device gave higher blood pressure readings than the arm device (difference in systolic/diastolic blood pressure: 6.3±17.3/3.7±11.8 mmHg, P<0.001). However, the arm device tended to give lower readings than the wrist device for high blood pressure values. The prevalence of blood pressure categories differed substantially depending on which device was used, 29% and 14% for blood pressure <120/80 mmHg (arm device versus wrist device, respectively), 30% and 33% for blood pressure 120–139/80–89 mmHg, 17% and 26% for blood pressure 140–159/90–99 mmHg, 12% and 13% for blood pressure 160–179/100–109 mmHg and 13% and 14% for blood pressure ≥180/110 mmHg.

Conclusions A large discrepancy in the estimated prevalence of blood pressure categories was observed using two different automatic measurement devices. This emphasizes that prevalence estimates based on automatic devices should be considered with caution.

aInstitute of Social and Preventive Medicine, University of Lausanne, Switzerland

bDivision of Hypertension and Vascular Medicine, Department of Medicine, CHUV, Lausanne, Switzerland

cMinistry of Regional Administration and Local Government, Dar es Salaam, Tanzania

Sponsorship: This research was financed by a grant of the Swiss National Science Foundation (No 32-51189.97).

Correspondence and requests for reprints to: Arnaud Chiolero, MD, Institute of Social and Preventive Medicine, 17, rue du Bugnon, 1005 Lausanne, Switzerland

Tel: +41 (0) 21 314 72 72; fax: +41 (0) 21 314 73 73;


Conflict of interest: None.

Received 04 November 2003 Revised 27 January 2004 Accepted 30 January 2004

© 2004 Lippincott Williams & Wilkins, Inc.