Comparison of the Average of Five Readings with Averages from Fewer Readings for Automated Oscillometric Blood Pressure Measurement in an Outpatient Clinic

BACKGROUND AND OBJECTIVES The optimal number of consecutive measurements for obtaining an average blood pressure (BP) reading in clinical practice is yet undefined by research. This study aimed to compare readings obtained from an average of 5 with averages of 2, 3, and 4, sequential measurements. SUBJECTS AND METHODS Using an automated oscillometric device (BpTRU), BP measurement was conducted on 410 consenting adults attending a general outpatients clinic. Comparison of an average of 5 readings with averages of 2, 3, and 4 readings involved evaluation of correlations, Bland-Altman analysis, comparison of means and distribution of readings, and determination of the proportion of differences between compared readings which were clinically non-significant. RESULTS 397 (96.8%) sets of complete BP readings were suitable for analysis. Clinically non-significant differences (≤5 mm Hg) were found between at least 79.3% (n=315) and 96.5% (n=383) of compared systolic and diastolic readings, respectively. Bland-Altman's analysis revealed that the 95% limits of agreement for the differences between compared readings were approximately 2-4, 3-7, and 4-11 mm Hg for 2, 3 and 4 readings' systolic comparisons while those for diastolic comparisons were 2-3, 3-5, and 4-7 mm Hg, respectively. Statistically non-significant differences were observed in all comparisons of the distributions of readings that were classified as <140 mm Hg or ≥140 mm Hg and <90 mm Hg or ≥90 mm Hg for systolic and diastolic readings, respectively. Strong positive correlations were found between 5 average readings and each of 2, 3, and 4 average readings, respectively. CONCLUSION An average of 5 readings may be excessive for routine BP measurement using this device at outpatient clinics.