Echocardiograms are largely performed for suspicion of infective endocarditis (IE), and usually have a poor diagnosis yield, especially if there is a low clinical probability. The ESC Guidelines of 2009 and 2015 proposed very large criteria that define clinical suspicion of IE and justify an echocardiogram request. We studied if the patients referred for suspicion of IE had clinical criteria who respected the ESC Guidelines for screening echocardiography, and we tried to identify the factors associated with the diagnosis of IE.

We retrospectively examined all the transthoracic echocardiogram (TTE) and transoesophageal echocardiogram (TOE), performed for suspicion of IE between January 2017 and December 2017 in the University Hospital of Nîmes, France.

Three hundred and fifty patients were referred for a suspicion of IE. Two hundred and eighty three (80,9%) TTEs and 67 (19,1%) TOEs were performed in first intention. A second exam was performed in 99 (28,3%) patients: a TTE for 56patients and a TOE for 43patients. For 61 (17,4%) patients, the echocardiogram request was not appropriate according to referral of the ESC Guidelines, with no sufficient clinical criteria that may allow to suspect an IE. None of these 61patients was diagnosed with an IE. IE was finally diagnosed in 51 (14,6%) patients. The significant independent factors associated with the diagnosis of IE were: positive blood culture (OR=26,6 IC95%=[3,6–195,9] P=0,001), embolic phenomena (OR=2,4 IC95%=[1,3–4,5] P=0,005), previous valvular heart disease (OR=2,7 IC95%=[1,1–6,7] P=0,003) and sepsis in absence of an extra cardiac source of infection (OR=12,6 IC95%=[4,4–35,9] P<0,001) (Fig. 1).

One fifth of the patients referred for an echocardiogram had no sufficient clinical criteria to suspect an IE according to the referral of the ESC Guidelines, and none of them was diagnosed with an IE. The use of these criteria for screening echocardiography for suspicion of IE may avoid many unnecessary echocardiograms.