The aim of the study is to demonstrate if using one procalcitonin (PCT)-guided rule of stop of antimicrobials, the incidence of infections by C.difficile and by Multi-Drug-Resistant (MDR) bacteria during the next six months may be significantly decreased.

Early administration of antimicrobials remains the mainstay of treatment of severe infections. Current guidelines of management of severe sepsis suggest that initial therapy of a patient should be reviewed after 48 to 72 hours. At that stage some patients are doing well, whereas others fail to respond. When microbiology cultures of biological specimens fail to provide information for the microbial cause of an infection and susceptibilities to antimicrobials, antimicrobial stewardship relies on the use of biomarkers and mainly procalcitonin (PCT). Data so far, suggest that early changes of serum PCT can inform about the prognosis of the septic patient, with greater values reflecting a worse outcome and higher mortality and that serial measurements within 48-72 hours provide adequate information of the appropriateness of the administered antimicrobials. Moreover the use of a procalcitonin guided-treatment in surgical as well as in non-surgical critically-ill patients, is seen to be non-inferior to the standard antibiotic approach and leads to a shorter antibiotic exposure, having possible beneficial effect on reducing microbial resistance and therapy costs.

In the largest study conducted so far, de Jong et al showed that PCT-guided stop of treatment was not only safe compared with standard of care antibiotic duration, but also led to a better outcome i.e. significant decrease of both 28-day and 1-year mortality. The results of this study are a major contribution in the field of critical care since they prove for the first time that PCT guidance of antimicrobial treatment allows not only proper antimicrobial stewardship but it is also associated with survival benefit. However, de Jong et al did not provide findings to explain the underlying mechanism of survival benefit. As a rule critically ill patients run two major risks coming from the long-term administration of antimicrobials; the first is infections by Clostridium difficile coming from the ecological damage of gut flora and the second is the risk of infections by multidrug-resistant (MDR) bacteria colonizing the gut. MDR is emerging after the ecological pressure of broad-spectrum antimicrobial usually administered to the critically ill patient.

Inclusion Criteria:

• Male or female
• In case of women, unwillingness to remain pregnant during the study period.
• Age more than or equal to 18 years
• Sequential Organ Failure Assessment (SOFA) score more than or equal to 2 points for patients admitted in the emergencies and with a more than or equal to a 2-point increase of admission SOFA score for hospitalized patients.
• Presence of one of the following infections: community-acquired pneumonia, hospital-acquired pneumonia, ventilator-associated pneumonia, bacteremia and acute pyelonephritis. Any infection with onset more than 48 hours post hospital admission is considered one hospital-acquired infection.

Exclusion Criteria:

• Failure to obtain written consent to participate
• Patients in pregnancy or breastfeeding. Women of child-bearing potential will be screened by a urine pregnancy test before inclusion in the study
• Patients receiving prolonged antibiotic therapies ( e.g. endocarditis, implantable device-associated infection, cerebral/hepatic abscess, osteomyelitis, meningitis)
• Patients with severe infections due to viruses or parasites (e.g. Dengue, Toxoplasma gondii, Plasmodium spp.)
• Patients infected with Mycobacterium tuberculosis.