Accuracy of blood pressure measured by Korotkoff versus oscillometric devices: a comparison using intra-aortic BP as reference
摘要
This study compared the accuracy and agreement of an automated Korotkoff device (Hanvon FF690) and an oscillometric device (Omron HBP-1320T3) in two parts.
MethodsIn the first part, 60 patients undergoing coronary angiography were enrolled. Two non-invasive devices were applied in a randomized alternating order, with invasive aortic blood pressure (BP) serving as reference. In the second part, 134 outpatients were enrolled; two oscillometric BP monitors were first used to simultaneously measure and quantify the baseline inter-arm difference (IAD) for each patient. Subsequently, the two study devices were randomly assigned to two groups for simultaneous measurement, and the data were corrected using the baseline IAD. Paired tests and Bland–Altman analysis were used to assess accuracy and agreement, and linear regression was used to evaluate correlation.
ResultsIn the accuracy assessment, both non-invasive devices significantly overestimated invasive aortic systolic blood pressure (SBP). The oscillometric device overestimated SBP by a larger margin than the automated Korotkoff device (mean overestimation 14.8 ± 20.6 mmHg vs. 9.9 ± 18.6 mmHg, P < 0.001). For diastolic blood pressure (DBP), both methods showed no systematic bias against the invasive reference standard. In the clinical agreement assessment, the oscillometric device gave significantly higher SBP and DBP readings than the Korotkoff device (mean differences: SBP 3.34 ± 8.81 mmHg, DBP 2.87 ± 6.34 mmHg, both P < 0.001). Stratified analysis showed that the between-device difference for SBP remained stable across BP levels, whereas the DBP measurement difference tended to narrow at higher baseline pressures.
ConclusionBoth the oscillometric and Korotkoff sound methods overestimated SBP by approximately 10–15 mmHg. Comparatively, the oscillometric SBP measurements remained 3–5 mmHg higher than those obtained via the Korotkoff sound method, with both approaches demonstrating substantial individual variability.