We aim to study whether the use of inhaled antibiotics may have an impact on the incidence of tracheobronchitis and pneumonia associated with the use of mechanical ventilation. The investigators hypothesize that this could be possible by decreasing the population of gram negative germs that colonize the lumen of tracheal tube and upper airways.

INTRODUCTION Gram-negative bacteria have recently become endemic in many ICUs around the world. Ventilator associated pneumonia (VAP) is relatively common and its parenteral treatment often leads to low and sub-therapeutic concentrations of antibiotics in the lung. Ventilator associated tracheobronchitis (VAT) is characterized by fever, purulent sputum, positive cultures of tracheal aspirates secretions and the absence of new pulmonary infiltrates on CXR. This entity is of particular interest because it is an intermediate step in the pathogenesis of VAP (between colonization and VAP itself) and prolongs mechanical ventilatory support. Inhaled antibiotics achieve high (both peak and trough) concentrations in respiratory secretions. Endobronchial antibiotics have been given to non-cystic fibrosis patients for prevention of VAP. Colistimethate sodium (colistin) given by instillation decreased the incidence of Gram-negative pneumonia and did not lead to the development of colistin resistant bacteria.

AIM

The study is a prospective open label randomized study which aims to clarify the effect of colistin to:

1. The incidence of VAP The incidence of VAT
2. Mechanical ventilation free days
3. Incidence of multidrug resistant bacteria in tracheal aspirates cultures

METHODS Settings: The study will be performed in the ICU of the University of Larissa (12 beds).

Entry criteria: All intubated and tracheostomised patients, older than 18 years old with duration of intubation of 48 hours.

Exclusion criteria: 1) Purulent sputum within 48 hours from admission. 2) Pneumonia within 48 hours of intubation 3) New chest X ray infiltrate, 4) Advanced COPD that led to intubation 5) Pregnancy 6) Allergy to colistin 7) Resistance of the bacterial strain to colistin.

Tools: Randomization of the patients into two groups: the first group will receive nebulized colistin and the other no treatment. The initial dose is 1000000 units and it will be doubled after measurement of the drug concentration in tracheal secretions. The drug will start on admission and stop after 10 days.

All patients will be given the respiratory bundle measures (semi-recumbent position, daily interruption of sedation) as part of the usual daily practice of our ICU except otherwise indicated.

The severity of patients will be estimated by APACHE II score on admission and SOFA score thrice weekly. CPIS score as a screening tool for pneumonia will be estimated as well thrice weekly.

Tracheal aspirates cultures on admission and thrice weekly thereafter for the first two weeks and twice weekly thereafter.

The patients will be followed for thirty days to measure outcome.

Inclusion Criteria:

• All intubated and tracheostomized patients, older than 18 years old with duration of intubation of 48 hours.

Exclusion Criteria:

• Purulent sputum within 48 hours from admission
• Pneumonia within 48 hours of intubation
• New chest X ray infiltrate
• Advanced COPD that led to intubation
• Pregnancy
• Allergy to colistin
• Resistance of the bacterial strain to colistin