To the Editor—Reply to Karakonstantis, Zervou, Rasmussen and coworkers for reading our article and for their cogent comments [1–3]. Certainly, an always challenging issue is how to apply clinical research to daily clinical practice. This concern seems to be even more critical in the prediction models’ scope, either for diagnosis or prognosis [4]. Therefore, updated, wide, and independent validation of any predictive model is just one step in a continuously evolving process. Karakonstantis et al asked the very important question of how sure are we about the use of the VIRSTA score in clinical practice [1]. High discordance exists in selection criteria for echocardiography in patients with Staphylococcus aureus bacteremia (SAB), even in series with expert clinicians [5]. Risk stratification using scores to limit the use of echocardiography in low-risk patients with SAB is an innovative approach and a research gap...