RESUMO
Background: Atrial fibrillation (AF) has affected millions of adults in the world. It is important to monitor and manage blood pressure (BP) in AF patients. The accuracy of BP monitoring in AF patients with noninvasive methods remains questionable, however. Objectives: To compare the accuracy of different noninvasive BP devices (oscillographic sphygmomanometer and pulse wave device) for BP measurement in elderly patients with AF, with a mercury sphygmomanometer as a reference. Design: This study was an observational study. Methods: Patients with AF from the inpatient department of cardiology were included from 1 January to 31 December 2020. BP measurements were performed by two trained nurses using a tee junction connection on the cuff to connect three sphygmomanometers. The Bland-Altman plot analysis was conducted to compare the agreement of BP measurements. We also compared the agreement of BP measurements through metrics such as accuracy, bias, and precision. Results: A total of 202 patients (54.5% female) were included. The Bland-Altman plot analysis showed that the lower and upper limits of agreement (LoAs) of pulse wave/reference were similar to the predefined acceptable clinical limits (10/5 mmHg). The bias and precision in both systolic and diastolic BP were significantly less in pulse wave/reference (a bias of 1.8 and 0.77 mmHg and a precision of 5.20 and 4.66 mmHg, respectively), with corresponding higher accuracy readings (98.51% for P10 in systolic BP and 85.64% for P5 in diastolic BP). Conclusion: A novel noninvasive sphygmomanometer - pulse wave device has a good concordance with a mercury sphygmomanometer in BP monitoring, and may be applicable to perform BP measurements in the elderly with AF.