1021-55 Is There Incremental Value of Echocardiography Over Clinical Information in Unselected Patients with Suspected Endocarditis? A Prospective Study

Abstract

Although both transthoracic (TTE) and transesophageal (TEE) echo can detect vegetations in patients with bacterial endocarditis (BE), their incremental value over the available clinical information has not been examined. Accordingly, we performed a prospective study in 105 patients (59 men, 46 women, age range of 16–88 yrs) with suspected BE in whom TIE had been requested. Thirty-four of these patients had known valvular disease, 19 had prosthetic valves, and 53 had other sources of infection. In addition to TTE, TEE was also performed in all patients. Twelve patients (11%) had technically inadequate TTE. Predefined clinical variables were used to classify the patients into low, medium, or high probability of having BE on clinical grounds. TTE and TEE studies were read separately without any clinical information and the patients were again classified into low, medium, or high probability of BE based on the echo findings alone. Also, of the 67 patients with low clinical probability of BE, the majority (79% by TTE and 87% by TEE) were also considered to have low probability by echo; only 6% were considered high probability. In patients with medium clinical probability of BE, the majority (83% by TTE and 100% by TEE) were reclassified based on echo findings. In patients with high clinical probability of BE, a significant portion (63% by TTE and 50% by TEE) were also reclassified based on echo findings. There was concordance between TTE and TEE in 86% of cases. The greatest difference was noted when TTE showed medium probability of BE: 8 of these 10 studies (80%) were reclassified based on TEE. We conclude that in patients with low clinical probability of BE, echo offers no incremental value, In patients with medium to high clinical probability of BE, echo offers incremental value. In addition, TEE is indicated in the diagnostic work-up only in patients with technically inadequate TTE studies and in those who show medium probability of BE after TTE; these constitute <10% of patients with suspected BE in whom echo is requested.