Article • Fighting drug resistance

Why antimicrobial stewardship must not stop at the ICU door

Antimicrobial stewardship is essential to curbing the growing threat of drug-resistant infections – yet its principles are too often abandoned the moment a patient leaves the intensive care unit (ICU). At the Critical Care Symposium in Manchester, Professor Marc Leone outlined why consistent antibiotic management across the entire hospital stay is vital for patient outcomes.

By Mark Nicholls

In the ICU, antibiotic consumption is high by necessity. Around 75% of patients receive antibiotics on the day of admission, and every course of treatment carries the risk of fuelling resistance. 'Each time you use antibiotics, there is some resistance growing; that is the interaction which is difficult to manage in routine,' Leone said, underlining the importance of antimicrobial stewardship (AMS).

The severity of the patient's condition dictates the approach. In cases of septic shock, blood cultures must be collected and antibiotic therapy started without delay. Where the probability of sepsis is low, however, there is room to pause and confirm whether an infection is actually present. 'If my patient is in septic shock, I should start immediately, though if the probability of sepsis is low, I can wait and add some examinations to confirm the infection or that there is no infection,' he explained.

Getting the dose and delivery right

Once an antibiotic has been selected, pharmacokinetic and pharmacodynamic protocols guide dosage. Leone highlighted continuous infusion of beta-lactams – commonly used in the ICU setting – as the preferred delivery method after an initial bolus. The target is a concentration of four to ten times the minimum inhibitory concentration (MIC).

'Now, we prefer continuous infusion after the first bolus by the fact that the first bolus, or even during the first 24 hours, the dose is independent of kidney function – with adaptation to each patient depending on renal and liver function,' he said. 'We prefer short treatment every time as source of infection is controlled.'

When the source of infection can be controlled, treatment duration should not exceed five to seven days. If the source cannot be controlled, longer courses may be necessary.

Sometimes you are fighting to reduce use of antibiotics during the ICU stage but as the patient is discharged into a surgical or other unit, broad spectrum treatment is sometimes initiated and all the efforts you made during the ICU stage are destroyed
Marc Leone

A central concern for Leone is what happens once a patient is discharged from the ICU to a surgical or medical ward. Too often, he warned, broad-spectrum antibiotics are restarted – undoing the careful stewardship efforts of the ICU team. 'Sometimes you are fighting to reduce use of antibiotics during the ICU stage but as the patient is discharged into a surgical or other unit, broad spectrum treatment is sometimes initiated and all the efforts you made during the ICU stage are destroyed; antimicrobial stewardship should not stop at the end of the ICU.'

He advocates establishing dedicated AMS teams that monitor treatment throughout the entire hospitalisation journey – and even after discharge, whether the patient goes home or to a rehabilitation unit.

The diagnostic challenge at the bedside

One of the most persistent challenges in the ICU is distinguishing infection from inflammation. 'At the bedside, the first challenge is to identify the patient with an infection and not inflammation and that is a real challenge because we do not have good biomarkers for this,' Leone said.

He pointed to C-reactive protein (CRP) and procalcitonin (PCT) as commonly used biomarkers, but cautioned that neither is ideal for diagnosing infection in critical care. CRP is sensitive but non-specific, rising in response to inflammation regardless of its cause. PCT, while more specific for bacterial infections, should not be relied upon for initial diagnosis either – though it can be useful for guiding early cessation of antibiotic treatment based on its kinetics. In the future, Leone expects that combined analysis of a panel of biomarkers will improve diagnostic accuracy.

Identifying the pathogen – and knowing when to stop

Equally important is the rapid identification of the bacteria responsible for the infection. Blood cultures and respiratory secretion samples remain the standard approach, but rapid diagnostic tests are now available that can deliver accurate results far more quickly. These tests are particularly valuable for healthcare-associated infections and are recommended in the latest 2026 Surviving Sepsis Campaign guidelines.¹

Those guidelines also strongly recommend active de-escalation of antimicrobial therapy – narrowing the spectrum once the pathogen has been identified. Yet Leone acknowledged that this remains difficult in practice. 'As you identify your pathogen, you can focus on a spectrum of treatment on this single bacteria though that process of de-escalation is not so easy in real life,' he said.

He pointed to the DIANA study,² a prospective multicentre investigation across 28 countries, which found that de-escalation was performed in only 16% of patients – a figure he described as 'frightening'. The final challenge, he added, is knowing when to stop treatment altogether.

Profile:

Marc Leone is Professor of Anaesthesiology and Critical Care Medicine at Aix-Marseille University and Chair of the Department of Anaesthesiology and Critical Care Medicine at Nord University Hospital in Marseille, France. His research focuses on sepsis, antimicrobial stewardship, biomarkers, and vasopressor use. He is a member of the European Society of Intensive Care Medicine (ESICM) and the French Society of Anaesthesia and Intensive Care Medicine (SFAR).

References:

05.08.2026