The measurement of blood pressure (BP) in atrial fibrillation (AF) is challenging and current guidelines recommend the use of the auscultatory method. The use of automated BP devices in AF is still debated. The objective was to analyze the reliability and the reproducibility of three different oscillometric BP devices in patients with AF and to assess the accuracy of two arm oscillometric BP devices in comparison with invasive blood pressure measurements (IBP).Design and method:
Forty-eight hospitalized patients with AF were randomized to undergo 10 consecutive BP measurements with either Omron R7™ (OR7, wrist) and HEM907™ (HEM, arm) or Omron R7™ and Watch BP home™ (WBP, arm). Measurements were stopped if a device failed to measure BP two times consecutively. For each device, reliability and reproducibility were assessed by within-subject variation and intraclass correlation coefficient (ICC). In another group of 10 patients with AF hospitalized in the intensive care unit, 10 consecutive measurements with IBP and HEM or IBP and WBP were performed, and results were compared using Bland Altman and Pearson correlation coefficient.Results:
OR7 was unable to measure BP in all patients. HEM provided 10 consecutive measurements in 35% of the patients, compared to 88% for the WBP (P < 0.01). The ICC between the arm devices was better for the systolic (HEM: 0.94, WBP 0.92) than for the diastolic (HEM: 0.77. WBP: 0.79) BP. HEM values showed a good correlation with IBP (r = 0.92 for both systolic and diastolic BP), while WBP values showed lower correlations (r = 0.85 for systolic and r = 0.52 for diastolic). Compared to IBP, the limits of agreement for systolic and diastolic BP were quite wide for both devices. However, when individual systolic and diastolic BP measures were averaged, the agreement improved except for the HEM systolic BP (table 1).Conclusions:
The wrist device cannot be recommended for clinical use in patients with AF. Both arm devices are reliable in patients with AF when they manage to measure BP; however, their accuracy seems better for systolic than for diastolic BP. The repetition of measurements improves their accuracy compared to IBP.