The study aims to compare the efficacy and safety of an absolute procalcitonin (PCT) value-guided antibiotic initiation protocol and a protocol using the kinetics of PCT (the difference between the actual and the previous day value) in hemodynamically stable critically ill patients with suspected new-onset infection on admission or during ICU stay.

The main question it aims to answer:

• Can the investigators decrease the number of unnecessary AB therapies using the kinetics of PCT insted of using absolute PCT values?
• Is it safe to use PCT kinetics together with the clinical picture to guide AB initiation? AB therapy will be initiated according to predefined PCT protocols (Kinetics and Absolute Group). After 72 hours of treatment, an independent multidisciplinary team (infectologist, microbiologist and intensivist) will decide about the necessity of the treatment with all the relevant results in hand.

Appropriate antibiotic (AB) therapy is still a big challenge in intensive care units today. More than 50% of our patients are considered potentially infected, and infection alone can double mortality; however, infection is later confirmed in less than 60% of patients admitted with the initial diagnosis of sepsis. According to previous studies, at least every fifth patient in intensive care units receives unnecessary antibiotics (ABs), leading to the well-known adverse effects of ABs without any benefit. The emergence of AB resistance has been associated with 700,000 avoidable deaths in 2014, and WHO estimates that it will contribute to the death of 10 million people by 2050.

Procalcitonin (PCT) is one of the most studied inflammatory biomarkers. Several randomized controlled trials (RCTs) and their meta-analyses concluded that PCT-guided antibiotic treatment could reduce the length of AB therapy without adverse effects; moreover, it may be associated with reduced 28-day mortality, which was also confirmed by our study in 2023. Accordingly, the current sepsis guideline recommends the combined evaluation of the clinical picture and PCT when stopping AB therapy; however, it contains a weak recommendation against using PCT when starting the treatment. The latter proposition is based on three studies with a significant number of surgical patients (around 40%) in two of them, in whom the applied PCT cut-off (0.5-1 ng/ml) presumably was too low, supported by several previous studies. Therefore, the overuse of ABs was more or less hardcoded into the protocol.

Kinetics of an inflammatory biomarker can carry much more information about the host response to infection instead of a single value. In a prospective observational study by Trasy et al., early PCT kinetics (PCT change between the day of suspected infection and the previous 24 hours) predicted infection, while PCT did not change in the group of patients in whom the infection was later ruled out. Tsangaris and his colleagues measured PCT daily in critically ill patients. The investigators noted a two-fold increase in PCT levels between the day of fever onset and the previous day if the patients had an infection, while there was no difference in the PCT values if the participants had no infection.

Based on these findings, the investigators created a protocol to reduce the number of unnecessary antibiotic therapies using the kinetics of PCT.

Inclusion Criteria:

• Adult (18 years < ) non-surgical, surgical, or trauma patients
• Suspected new-onset infection on admission or during ICU stay
• The source of infection is known or highly suspected, and source control has been implemented if needed (i.e., removal of an infected device (e.g., central line, endoprosthesis)
• Two PCT values are available - one on the day of suspicion of infection and one 24±4 hours earlier.
• Microbiology sampling has to be performed (according to all presumed sources - blood culture -aerobic and anaerobic, lower respiratory tract sample (tracheal aspirate/bronchoalveolar lavage), urine, etc.).
• Written informed consent of the patient (or legal guardian if the patient cannot provide consent)

Exclusion Criteria:

• Septic shock (hypotension requiring vasopressor therapy to maintain mean blood pressure of 65 mmHg or greater and having a serum lactate level greater than 2 mmol/L after adequate fluid resuscitation)
• Infections for which long-term antibiotic treatment is strongly recommended (e.g., infective endocarditis, osteoarticular infections, chronic prostatitis, tuberculosis)
• Infections related to primary surgical intervention and adequate source control cannot be guaranteed (e.g., fecal peritonitis, pancreatic necrosectomy, infective necrotizing fascitis - i.e., Fournier's gangrene),
• Indisputable infections (e.g., hepatic abscess, empyema)
• Poor chance of survival (i.e., expected ICU stay less than 24 hours or initial Acute Physiology and Health Evaluation Score II (APACHE II) >30)
• Admissions after cardiopulmonary resuscitation
• Severe immunosuppression other than steroid use
• stem-cell transplant recipients
• solid organ transplant patients
• HIV infection with a CD4 count of less than 200 cells/mm3
• Neutropenia with less than 500 neutrophils/mm3
• Patients on ABs within 72 hours before inclusion
• Patients in pregnancy or breastfeeding. Women of childbearing age will be screened by a urine pregnancy test before inclusion in the study.

All data obtained through the study may be provided to researchers with an interest in optimizing antibiotic therapy of critically ill patients. Data and samples will be coded if shared with no protected health information included.

Zsolt Molnár will receive invitations/requests to share individual participant data, and will consider if the request: 1) explicitly serves public benefit, 2) complies with local ethical and data security regulations of the requesting and providing institutions, 3) lays out the existing or planned legal and technical infrastructure for data transfer, 4) provides a publicly available protocol for planned statistical analyses.

Nurses