Ambulatory blood pressure monitoring (ABPM) is considered the gold-standard for the diagnosis of hypertension and assessment of cardiovascular risk. This study aimed to compare the reliability of a cuff-less blood pressure (BP) method based on pulse transit time (PTT) versus oscillometric ABPM considering limitations of cuff based BP measurement such as movement artefacts, arrhythmia and arousals during sleep caused by inflation of the cuff.Design and method:
Systolic (SBP) and diastolic (DBP) blood pressure were measured in 30 adults (9 women, mean age 65.5 ± 9.3 years) using an oscillometric 24 hour ABPM (boso TM-2430 PC2). Measurement intervals were 15 min during daytime (6–22 h) and 30 min during night-time (22–6 h). Simultaneously, BP was recorded beat-to-beat and non-invasively based on PTT (SOMNOtouch NIBP, SOMNOmedics GmbH) on the contralateral arm. In addition we recorded a 3-channel ECG, motoric activity, finger plethysmogram, oxygen saturation and cuff pressure.Results:
In total 1293 measurements were analysed. The correlation coefficient for SBP was r = 0.77 (DBP: r = 0.53). The limits of agreement in Bland-Altman plot were +28 and −25 mmHg, with a mean difference of 1.6 mmHg for SBP (DBP: +30 and -26 mmHg; mean difference of 1.9 mmHg). 29% of all measurements were affected by the following artefacts: movement (18%), arrhythmia (5%) and arousal reactions (6%, corresponding to 18% of all nocturnal BP measurements). After elimination of all measurements influenced by cuff related artefacts the correlation of the remaining values improved significantly (n = 921; SBP: r = 0.89; DBP: r = 0.7). The limits of agreement in Bland-Altman plot were +18 and -15 mmHg with a mean difference of 1.4 mmHg for SBP (DBP: +14 and -13 mmHg; mean difference of 0.3 mmHg).Conclusions:
Our results imply that the cuff method is highly affected by movement artefacts, arrhythmia and arousal reactions during sleep. By exclusion of these artefact related measurements the correlation and agreement between ABPM and PTT method increased considerably (SBP: r = 0.89; DBP: r = 0.7). This indicates that oscillometric ABPM is affected by various artefacts which limit the reliability of cuff based BP measurement.