Accuracy of abbreviated protocols for unattended automated office blood pressure measurements, a retrospective study

Abstract

Background Blood pressure measurement (BPM) is one of the most often performed procedures in clinical practice, but especially office BPM is prone to errors. Unattended automated office BPM (AOBPM) is somewhat standardised and observer-independent, but time and space consuming. We aimed to assess whether an AOBPM protocol can be abbreviated without losing accuracy. Design In our retrospective single centre study, we used all AOBPM (AOBPM protocol of the SPRINT study), collected over 14 months. Three sequential BPM (after 5 minutes of rest, spaced 2 minutes) were automatically recorded with the patient alone in a quiet room resulting in three systolic and diastolic values. We compared the mean of all three (RefProt) with the mean of the first two (ShortProtA) and the single first BPM (ShortProtB). Results We analysed 413 AOBPM sets from 210 patients. Mean age was 52±16 years. Mean values for RefProt were 128.3/81.3 mmHg, for ShortProtA 128.4/81.4 mmHg, for ShortProtB 128.8/81.4 mmHg. Mean difference and limits of agreement for RefProt vs. ShortProtA and ShortProtB were -0.1±4.2/-0.1±2.8 mmHg and -0.5±8.1/-0.1±5.3 mmHg, respectively. With ShortProtA, 83% of systolic and 92% of diastolic measurements were within 2 mmHg from RefProt (67/82% for ShortProtB). ShortProtA or ShortProtB led to no significant hypertensive reclassifications in comparison to RefProt (p-values 0.774/1.000/1.000/0.556). Conclusion Based on our results differences between the RefProt and ShortProtA are minimal and within acceptable limits of agreement. Therefore, the automated procedure may be shorted from 3 to 2 measurements, but a single measurement is insufficient.