Accuracy of oscillometric blood pressure measurement using a Cardell Touch multiparameter monitor in anesthetized pigs

To assess the accuracy of an oscillometric monitor in anesthetized pigs and its ability to track changes in mean arterial pressure (MAP) and to detect arterial hypotension and hypertension in reference to direct measurements. Mean arterial pressure was measured simultaneously from a catheter in a femoral artery and with an oscillometric cuff placed over the metatarsus in 9 anesthetized pigs (∼6 months old, 35 to 55 kg). Pigs were subjected to maneuvers to alter MAP. Paired values for invasive and noninvasive MAP (iMAP and NiMAP) were recorded every 2 minutes. Bland-Altman plots with bias, limits of agreement, and percentage error were constructed using each pair and the average of 5 consecutive values. Concordance was calculated using changes in MAP at 20-minute intervals. Receiver operating characteristics (ROC) curves were constructed to test NiMAP for detection of hypotension (< 70 mm Hg) and hypertension (≥ 120 mm Hg). Bias of NiMAP was -8.59 mm Hg for consecutive pairs and -8.85 mm Hg for averaged pairs, relative to iMAP. Limits of agreement and percentage error were reduced for averages (19%) over individual pairs (26%). Concordance was 82%. The area under the ROC curve for detecting hypotension with NiMAP was 0.936, with a best cutoff value of 63 mm Hg NiMAP. The area under the ROC curve for hypertension was 0.940, with a best cutoff value of 101 mm Hg NiMAP. Averaging several consecutive values improves the accuracy of NiMAP measurements. This device correctly tracked changes in MAP approximately 80% of the time and appears reliable for diagnosing arterial hypotension.