Background and aim of the study: Intraoperative transesophageal echocardiography (TEE) can be used to assess gradients after aortic valve replacement (AVR). However, altered cardiac output after weaning from cardiopulmonary bypass, dynamic changes in valve function early after surgery, and limitations in transducer alignment may compromise the ability to predict valve hemodynamics on follow up. The study aim was to compare gradients on intraoperative TEE immediately after AVR with gradients on transthoracic echocardiography (TTE) after two to four months.
Methods: Among 90 patients who underwent AVR between September 2002 and April 2003, 31 had both intraoperative ‘post-pump’ TEE, including transaortic gradients, and TTE at two to four months after surgery. Mean and peak transvalvular gradients were compared between time periods, and a receiver

operator curve (ROC) was constructed to test the ability to predict a mean gradient >15 mmHg (upper quartile of patients) on follow up.
Results: There was no discernable correlation between mean gradient on intraoperative imaging and at follow up (R2 = 0.09; p = 0.11); for peak gradient there was a statistically significant but extremely weak correlation between time points (R2 = 0.17; p = 0.02). A ROC revealed very poor prediction of high mean gradient on follow up based on mean gradient on intraoperative TEE (area under curve 0.63).
Conclusion: Gradients on intraoperative TEE immediately after AVR do not correlate with later gradients on TTE. The mean gradient on intraoperative TEE holds no useful predictive value for a high mean gradient at follow up.
The Journal of Heart Valve Disease 2004;13:881-886