PS-BPC04-9: IMPACT OF ATRIAL FIBRILLATION ON ACCURACY OF OSCILLOMETRIC BLOOD PRESSURE MONITORING IN CRITICALLY ILL PATIENTS FROM A LARGE REAL-WORLD DATABASE
Yuxi Li,Beining Wang,Fangfang Fan,Haoyu Weng,Long Zhang,Jie Jiang,Bo Zheng,Yan Zhang,Jianping Li
DOI: https://doi.org/10.1097/01.hjh.0000915812.01696.80
IF: 4.9
2023-01-01
Journal of Hypertension
Abstract:It is controversial to use oscillometric blood pressure (BP) monitors for BP measurement in atrial fibrillation (AF) patients due to the variation in stroke volume. Several recent studies revealed no significant difference in oscillometric and intra-arterial BP bias between AF and sinus rhythm (SR) patients, but the population sizes were limited. We performed a cross-sectional study based on an extensive real-world database to further demonstrate the impact of AF on the accuracy of oscillometric blood pressure monitoring in the intensive care unit setting. Adult patients with AF or SR records were enrolled from Medical Information Mart for Intensive Care-III (MIMIC-III) database. Concurrently-recorded non-invasive oscillometric BPs (NIBPs) and invasive intra-artery BPs (IBPs) were extracted and classified as AF or SR group according to the heart rhythm. Pairwise comparison was performed on NIBP/IBP bias between AF and SR circumstances. Bland-Altmann plots were used to assess bias and limits of agreement. A multivariable linear mixed-effect model was used to adjust confounders. Subgroup and interaction analyses were performed. 2335 patients (71.95 ± 11.23 years old, 60.90% of male) were included. The systolic (SBP), diastolic (DBP), and mean (MBP) NIBP/IBP biases were not clinically different between AF and SR circumstances (SBP: AF vs. SR, 0.66 ± 19.64 mmHg vs. 1.21 ± 21.96 mmHg, P = 0.002; DBP: -5.29 ± 12.03 vs. -5.17 ± 11.54, P = 0.124; MBP: -4.45 ± 12.31 vs. -4.19 ± 12.61, P = 0.014). After adjusting age, gender, heart rate, arterial BP, and vasopressor usage, the effect of heart rhythm on NIBP/IBP bias was still within ± 5 mmHg for SBP and DBP [adjusted effect on SBP bias: 3.32 mmHg (95% CI 2.89 to 3.74), P < 0.001; DBP: -0.89 (-1.17 to -0.60), P < 0.001], while the effect on MBP bias was not significant [adjusted effect 0.18 mmHg (-0.10 to 0.46), P = 0.200]. Atrial fibrillation would not significantly influence the accuracy of oscillometric monitoring in critically ill patients compared with sinus rhythm circumstance. Bland-Altmann plots comparing systolic (A), diastolic (B), and mean (C) noninvasive oscillometric blood pressures (NIBPs) to the gold standard of invasive arterial blood pressures (IBPs) in circumstances of atrial fibrillation and sinus rhythm. NIBP/IBP bias is represented as dots. Dotted lines provide mean BP bias and 95% limits of agreement (±2 × SD). Linear fitting curves are shown as dashed lines. A-fib, atrial fibrillation; Sinus, sinus rhythm; SBP, systolic blood pressure; DBP, diastolic blood pressure; MBP, mean blood pressure.
peripheral vascular disease