PS-C11-6: SIMULTANEOUS MEASUREMENT OF 24-HOUR AMBULATORY BLOOD PRESSURE IN BOTH ARMS REVEALS CLINICALLY RELEVANT DIFFERENCES: A WAKEUP-CALL FOR HYPERTENSION MANAGEMENT

Abstract

Objective: Cuff-based 24-hour ambulatory blood pressure measurement is the gold standard for hypertension diagnosis and treatment monitoring. The clinical value of ambulatory blood pressure (BP) measurement is ubiquitously accepted. This study wants to quantify the measurement discrepancies in simultaneous and synchronized 24-hour ambulatory BP measurements on both arms. Identifying and understanding measurement uncertainties could improve future hypertension management and originate true confidence in alternative, cuff-less systems with equal measurement precision. Methods: We conducted simultaneous 24-hour ambulatory blood pressure measurements on both arms in 52 healthy young adults. The broadly used and validated cuff-devices (BoSo TM-2430) were synchronized to start BP measurement within one second of each other. Further, we recorded, cuff-pressures of both cuff-based devices and the hydrostatic pressure differences between both cuffs. We then scored pressure curves for disturbances (e.g., pressure spikes during deflation) and were able to correct measurements for hydrostatic differences between the arms. This enabled us to not only quantify the extent of measurement uncertainty but to describe its origins and possible countermeasures. Results: Measured BP levels from left and right arm showed limits of agreement in the Bland Altman Plot of > 33 mmHg for systolic and > 30 mmHg for diastolic values over 24 h. Daytime values deviated more than nocturnal values (systolic: > 35 mmHg vs. > 27 mmHg systolic, and diastolic: > 32 mmHg vs. > 24 mmHg). Linear regression retrieved R2 = 0.39 (p < 0.001) for systolic and R2 = 0.20 (p < 0.001) for diastolic values. Exclusion of disturbed cuff pressure curves improved limits of agreement to > 23 mmHg for systolic and > 19mmHg for diastolic measurements. Subsequently, it improved correlation coefficients to R2 = 0.54 (p < 0.001) for systolic and R2 = 0.35 (p < 0.001) for diastolic values. Correction of hydrostatic differences between the arms had an negligible effect on limits of agreement and strength of correlation. Conclusion: Comparison of synchronized 24-hour BP measurements on both arms showed large measurement uncertainties with possible implications on clinical grading of hypertension and dipping pattern. Exclusion of disturbed cuff pressure curves, not detected by the conventional software of BP devices, mitigates the effect but still results in worrying discrepancies. These results are an imperative to further investigate the clinical implications of BP measurement uncertainty. It further inspires the courageous exploration of alternative means of BP measurement which might offer comparable measurement certainty.