Secondary Logo

Institutional members access full text with Ovid®

Discarding the first of three nurse-auscultatory or oscillometric blood pressure measurements does not improve the association of office blood pressure with ABPM

Graves, John W.a; Grossardt, Brandon R.b

doi: 10.1097/MBP.0b013e328337ce76
Analytical Methods and Statistical Analysis

Objective Guidelines for office blood pressure (BP) measurements recommend two or more readings at each visit. It has been suggested that discarding the first reading may improve the diagnostic utility of office readings. We compare two methods of office BP measurement to determine whether discarding the first BP affects their association with ambulatory blood pressure monitoring (ABPM) readings.

Methods We studied measurements in 313 mildly hypertensive patients (181 men, median age 51 years) at the end of the washout phase of a hypertension treatment trial. All patients had three auscultatory readings by a trained nurse using a mercury manometer before a 24-h ABPM and three automated readings (Omron 705 CP) to American Heart Association standards after ABPM. Daytime ABPM (09:00–21:00) of systolic BP and diastolic BP were compared with office methods using first only, second and third only, and all three readings. Associations were assessed using Bland–Altman plots and Pearson's correlations.

Results Bland–Altman plots revealed no differences between office methods and daytime ABPM regardless of whether the first reading was included or discarded. Results were similar for both systolic BP and diastolic BP. The Pearson correlations of office method readings with ABPM readings were lowest using the first measurement alone; however, discarding the first measurement did not improve the correlation compared with using all three readings.

Conclusion Discarding the first office BP reading does not change the association of office BP with ABPM or improve the correlation with ABPM values in hypertensive patients.

aDivision of Nephrology and Hypertension

bDivision of Biomedical Statistics and Informatics, Mayo Clinic College of Medicine, Rochester, Minnesota, USA

Correspondence to Dr John W. Graves, MD, FACP, Division of Nephrology and Hypertension, Mayo Clinic and Mayo Foundation, 200 First Street SW, Rochester, MN 55905, USA

Tel: +1 507 284 8712; fax: +1 507 281 1161; e-mail:

Received 7 July 2009 Revised 29 September 2009 Accepted 13 November 2009

This work was presented in part at the Annual Meeting of the European Society of Hypertension in June 2007.

© 2010 Lippincott Williams & Wilkins, Inc.