The tremendous development of Medicine with the advent of transplants, chemotherapies and immunotherapies has made our patient population more vulnerable to infections. For a long time, we could treat them with broad spectrum antimicrobials without the risk of “missing the target”, and adapt our treatment to the results of the microbiological analysis, when it was finally available. This approach has come and gone. The emergence and rapid dissemination of multidrug resistant (MDR) pathogens parallel to the difficulties encountered to provide new antimicrobials, has led today's clinicians in a difficult situation. They need to limit their usage of large spectrum antibiotics, despite being confronted to a higher risk of MDR pathogen. Antimicrobial stewardship (AMS) programs have been intensively developed during the last decades in hospitals, in an attempt to spare our “last-resource” antimicrobials and avoid the continuous spread of MDR bacteria. Microbiology has continually improved by providing clinicians with an accurate diagnosis of the pathogen and its susceptibility to antimicrobials. However, the long delay between sampling and results has limited its ability to influence the choice of antimicrobial in the first week, an unacceptable situation for AMS. Several revolutions took place in our microbiology labs in recent decades with the advent of mass spectrometry and automated bacterial and viral DNA detection systems that have significantly reduced the “time to results”. The line of research described in this thesis reflects a ongoing attempt to improve the “lead time” in our lab, by evaluating some of these new rapid diagnostic tests for their accuracy, reliability and potential impact on patient care. Those tests that convinced us based on these criteria were finally introduced inside our laboratory.