Assess the impact of a strategy combining respiratory mPCR and algorithm-based therapy developed using local epidemiology on the early optimization of initial antibiotic therapy for ventilator-associated pneumonia (VAP) (intervention), compared to a conventional strategy (control).

A bicentric, parallel-group, randomized controlled trial. The primary assessment criterion is the proportion of early optimized antibiotic therapy within 24 hours of respiratory sampling.

In both arms, for eligible patients with VAP, a deep respiratory sample collection is performed prior to inclusion. This collection is carried out either by mini bronchoalveolar lavage (BAL) via bronchoscopy or by blind mini BAL in alignment with the protocolized procedure. Initial antibiotic therapy is initiated after the deep sample has been taken according to the discretion of the attending clinician and in accordance with good practice recommendations. Both groups have their samples analyzed using conventional techniques. First, there is a direct examination, followed by culture and susceptibility testing, which is used as the gold standard technique for evaluating the primary endpoint.

In the intervention arm, in addition to conventional techniques, a broad-panel respiratory mPCR is performed before the 12th hour following achievement of the deep respiratory sample on the collected BAL. Once the mPCR results are received, the clinician adjusts the initial antibiotic therapy using algorithm-based therapy developed using local epidemiology in order to optimize treatment as early as possible.

In the control arm, the strategy is based on the clinician's choice in accordance with best practice recommendations without the combined use of respiratory mPCR and algorithm-based therapy. The deep respiratory sample is analyzed using conventional methods. Initial antibiotic treatment is therefore adapted by the clinician in charge of the patient according to departmental practices based on best practice recommendations, taking into account the results of the direct examination, culture, and susceptibility testing.

Inclusion Criteria:

• Adults (≥18 years) with VAP (mechanical ventilation and hospitalization ≥ 48 hours) and deep respiratory sample by mini BAL < 12 hours. The diagnosis of pneumonia includes two clinical criteria among fever (≥38.3°C), purulent sputum or aspiration, hyperleukocytosis (>12,000 WBC/mm³) or leukopenia (<4,000 WBC/mm³), hypoxemia, auscultatory signs in the affected area, and a newly-appeared parenchymal infiltrate
• Patient receiving initial probabilistic antibiotic therapy for VAP suspicion
• Informed consent or emergency procedure
• Patient affiliated with or beneficiary of a health insurance plan.

Non inclusion Criteria:

• Pregnancy
• Congenital immunodeficiency;
• HIV infection with the lymphocyte CD4 count below 200/mm3 or unknown in the last year;
• Acute hematologic malignancy;
• Neutropenia (<1 leucocyte/mL or < 0.5 neutrophil/mL);
• Immunosuppressive drugs within the previous 30 days, including anti-cancer - Chemotherapy and anti-rejection drugs for organ/bone marrow transplant
• Corticosteroids ≥ 20 mg/d of prednisone equivalent for more than 14 days
• Known allergy to beta-lactams
• Moribund patient or death expected from underlying disease during the current admission;
• Patient deprived of liberty or under legal protection measure;
• Participation in another interventional trial.

Exclusion criteria :

• mPCR non available

Outcome assessors (investigator-delegated evaluators, different from those who performed the procedure) will be blinded to treatment allocation.