Comparing Oscillometric and Manual Mean Arterial Pressure Measurements Among Children in the Emergency Department

OBJECTIVE: To evaluate agreement among three approaches to mean arterial pressure (MAP) measurement in children. METHODS: We performed a single-center prospective study of a convenience sample of children (<18 y) presenting to a pediatric emergency department. We compared MAP agreement using the intraclass correlation coefficient (ICC) measured in three ways: calculated via auscultation, calculated from oscillometric systolic and diastolic measurements, and measurement via oscillometry. RESULTS: We included 347 patients (median age 8 y, IQR: 4 to 13). The highest ICC was observed between calculated and device-reported oscillometric MAP measurements (0.83, 95% CI: 0.79-0.86), indicating good agreement. Agreement was poor between manual and device-reported oscillometric MAP (ICC: 0.33, 95% CI: 0.22-0.43) and between manual and calculated oscillometric MAP (ICC: 0.38, 95% CI: 0.20-0.52). The median difference between calculated oscillometric and device-reported oscillometric MAP was 1.0 mm Hg (IQR: 0.0 to 2.3), compared with -3.3 mm Hg (IQR: -8.3 to 2.0) between manual and device-reported oscillometric MAP and -4.0 mm Hg (IQR: -9.3 to 0.3) between manual and calculated oscillometric MAP. Bland-Altman plots suggested that manual MAP measurements were slightly lower than both device-reported oscillometric and calculated oscillometric MAP measurements. Agreement between calculated oscillometric and device-reported oscillometric MAP was highest among children <2 years and lowest among children 12 to <18 years. Agreement involving manual MAP remained poor across most age groups. Similar patterns were observed when stratified by weight and patient state. CONCLUSIONS: Calculated oscillometric MAP had good agreement with device-reported oscillometric MAP, while manual MAP had poorer agreement with either oscillometric measure. Further study is needed to determine the clinical implications of these differences across age, weight, and patient-state subgroups.