| Abstract: | To test the sensitivity and accuracy of pulsed Doppler ultrasonography to detect and quantitate left to right aortopulmonary shunt flow, an arterial allograft aortopulmonary anastomosis was constructed in nine adult mongrel dogs. Cardiac output and allograft flow were measured as the diameter of the allograft was varied. Piezoelectric crystals attached to the carotid artery and proximal descending aorta were energized with 20 MHz pulsed Doppler signals. Negative Doppler shift and negative Doppler shift/positive Doppler shift were calculated for seven dogs. All dogs exhibited negative Doppler shift in the carotid artery at zero allograft flow; five of the seven dogs exhibited a similar pattern in the descending aorta. Increasing negative Doppler shift was measured in all dogs from both sites as the allograft flow increased. Excellent linear correlation existed between allograft flow and negative Doppler shift and negative Doppler shift/positive Doppler shift for each dog from both sampling sites. However, marked interanimal variation in the slopes of the linear regression lines existed, making the composite linear correlation very poor. Detection of small left to right aortopulmonary shunting and single measurements to quantitate accurately left to right aortopulmonary shunting introduce errors due to intersubject variation. However, these results suggest that serial ultrasound measurements made over a short time can accurately predict changes in left to right aortopulmonary shunting. |
