• Sean Keane and Ignacio Martin-Loeches.
• Department of Anaesthesia and Critical Care Medicine, St. James's Hospital, Dublin 8, Ireland.
• Crit Care. 2019 Jun 14; 23 (Suppl 1): 134.

AbstractPatients admitted to the intensive care unit (ICU) often require invasive mechanical ventilation. Ventilator-associated lower respiratory tract infections (VA-LRTI), either ventilator-associated tracheobronchitis (VAT) or ventilator-associated pneumonia (VAP), are the most common complication among this patient cohort. VAT and VAP are currently diagnosed and treated as separate entities, viewed as binary disease elements despite an inherent subjectivity in distinguishing them clinically. This paper describes a new approach to pulmonary infections in critically ill patients. Our conjecture is that the host-pathogen interaction during mechanical ventilation determines a local compartmentalized or systemic de-compartmentalized response, based on host immunity and inflammation, and the pathogenic potential of the infecting organism. This compartmentalized or de-compartmentalized response establishes disease severity along a continuum of colonization, VAT or VAP. This change in approach is underpinned by the dissemination hypothesis, which acknowledges the role of immune and inflammatory systems in determining host response to pathogenic organisms in the lower respiratory tract. Those with intact immune and inflammatory pathways may limit infection to a compartmentalized VAT, while immunosuppressed mechanically ventilated patients are at greater risk of a de-compartmentalized VAP. Taking this model from the realm of theory to the bedside will require a greater understanding of inflammatory and immune pathways, and the development of novel disease-specific biomarkers and diagnostic techniques. Advances will lead to early initiation of optimal bespoke antimicrobial therapy, where the intensity and duration of therapy are tailored to clinical, immune and biomarker response. This approach will benefit towards a personalized treatment.

26 keywords...

hide…