Comparison of ultrasonic Doppler flow monitor, oscillometric, and direct arterial blood pressure measurements in ill dogs.

OBJECTIVE: To compare blood pressure measurements obtained via ultrasonic Doppler flow monitor (DOP) and 2 oscillometric noninvasive blood pressure monitors (CAR and PAS) to invasive blood pressure (IBP) in hospitalized, conscious dogs with a range of blood pressures.
DESIGN: Prospective clinical study.
SETTING: University teaching hospital. ANIMALS: Eleven client-owned dogs aged between 4 months and 11.5 years (median 6 y), and weighing between 5.8 and 37.5 kg (median 30.2 kg).
INTERVENTIONS: Blood pressure measurement.
MEASUREMENTS AND MAIN RESULTS: Three consecutive measurements of systolic, diastolic, and mean arterial pressure (MAP) were recorded for each of the 3 indirect devices (only systolic for DOP), along with concurrent IBP measurements. The data were categorized into 3 groups: hypotensive (direct MAP<80 mm Hg), normotensive (80 mm Hg<or=direct MAP>or=100 mm Hg), and hypertensive (direct MAP>100 mm Hg). Each indirect method was compared with the corresponding direct arterial pressure using the Bland-Altman method. Within the hypotensive group, each indirect method overestimated the corresponding IBP. Within the normotensive group all indirect systolic measurements and the PAS diastolic measurements underestimated the corresponding IBP. The remaining indirect measurements overestimated the corresponding IBP. Within the hypertensive group, DOP and CAR systolic measurements underestimated the corresponding IBP, and the remaining indirect measurements overestimated the corresponding IBP. In hypertensive dogs oscillometric systolic measurements were more accurate than MAP. In hypotensive dogs MAP measurements were more accurate than systolic measurements. All indirect measurements were most accurate in hypertensive dogs.
CONCLUSIONS: The noninvasive blood pressure monitors in our study did not meet the validation standards set in human medicine. However, CAR diastolic and MAP measurements within the normotensive group, CAR MAP measurements within the hypertensive group, and PAS diastolic measurements in all groups were close to these standards. All indirect measurements showed greater bias during hypotension. Precision was poorer for all indirect systolic measurements than for MAP.