For the detection and monitoring of hypertension, several consecutive blood pressure measurements are recommended to account for variability in BP. The rule is to perform three consecutive measurements after a rest period of a few minutes. The objective of the study was to look for the threshold at which the 3 consecutive measurements are no longer useful for evaluating BP.

In the PREDIC-HTA survey, consecutive taking of 3 measurements of BP in a seated position according to a standardized methodology was carried out using an automatic blood pressure monitor (BP5, iHealth, Paris) connected to a tablet and allowing the realization of a self-screening protocol for BP in volunteers. For each subject, the value of the first BP measurement was compared to the average of measures 2 and 3. The analysis focused on the difference between the “M1” and the “M2M3 average” with the search for a threshold of agreement. Two methods were used for threshold search.

In 6527 subjects, 3 BP measurements can be analyzed. BP is different and decreasing between measurements: 125.0/82.8 (M1) and 122.2/82.2 (M3). Bland-altman plot observed a positive regression line with an intersection at the value 0 observed for a SBP at 134mmHg and a DBP at 84mmHg. Regression to the mean analysis indicates that if M1 SBP is>135/85 then mean M2/M3 is −9.1/−2.2mmHg lower, and if M1 SBP is≤135/85 then mean M2/M3 is −0.8/−0.3mmHg lower.

When three BP measurements are performed with an automatic blood pressure monitor and in a standardized way, the first measurement is comparable to the average of the next two when the BP does not exceed the 135/85 values. When the first measure of BP does not exceed 135/85, there is no point in carrying out additional measurements. This result should be taken into account when the BP measurement is done by a health professional or a patient during self-measurement.