COPD is a common chronic disease. Its natural course is characterized by Acute exacerbations (AE). This may require hospitalization or even ICU/RESUSCITATION admission. The most common causes are respiratory distress with hypercapnic acidosis that requires mechanical ventilation (Invasive or non-invasive). Lower respiratory tract infections, bacteria and/or viruses are the main pathogenic factors of AE. The treatment of AECOPD is initially symptomatic treatment, combining bronchodilators, ventilatory support (oxygen therapy and/or mechanical ventilation) and respiratory physiotherapy. Systemic corticosteroid therapy is optional. When i) the sputum is purulent and ii) increased dyspnea and / or an increase in sputum volume is observed, antibiotic treatment is recommended for hospitalized patients. Antibiotic therapy is routinely recommended when mechanical ventilation is required.

During ICU/RESUSCITATION AECOPD, more than 85% of patients received antibiotic therapy, with a median duration of 8 to 9 days, and the benefit of antibiotic therapy is likely to be limited to infected patients. Suspected or documented lower respiratory tract bacteria, that is, 25% to 50% of patients. This will lead to overuse of antibiotics, which is a problem for patients and the community.

A personalized antibiotic strategy could limit this phenomenon, relying on multimodal methods, using aspect of sputum (clinical method), procalcitonin (PCT) (biological method) and the FilmArray ™ Pneumonia Panel extended panel multiplex respiratory PCR Plus (mPCR FA-PPP) (Biomérieux®) (microbiological approach).

The hypothesis of this study is that sputum appearance, procalcitonin (PCT) and the FilmArray ™ Pneumonia Panel Plus expanded panel multiplex respiratory PCR (mPCR FA-PPP) (Biomérieux®) could be used in combination , and their results integrated into a decision-making algorithm aimed at personalizing antibiotic therapy and guiding its early termination in patients admitted to ICU/RESUSCITATION due to acute exacerbation of chronic obstructive pulmonary disease (AECOPD) to the main benefit of antibiotic savings, and without additional risk to patient safety.

Inclusion (D0_H0) is performed in ICU/RESUSCITATION. The interval between admission to the hospital and admission to ICU/RESUSCITATION must be maximum 72 hours. Conventional microbiological investigations are left at the discretion of the physicians, and may include blood cultures, L. pneumophila and S. pneumoniae antigens. Usual biology includes procalcitonin measurement. Empirical antimicrobial therapy must be started as soon as possible after inclusion.

Randomization is performed immediately after the inclusion. In the intervention arm, a broad panel respiratory mPCR FA-PPP is performed on respiratory tract sample (tracheal aspirate, BAL or sputum), collected 12 hours after inclusion. An algorithm of early antibiotic adaptation and discontinuation, based on the microbiological results, including the mPCR FA-PPP results, and the procalcitonin values and kinetics and also aspect of sputum will be used. This algorithm will be applied as soon as possible after inclusion, and repeated day after day until D7.

In the control arm, the antimicrobial therapy is left at the discretion of the physicians, as in usual practice.

Evaluation criteria are collected at hospital discharge or at D28, and D90. The vital status may be obtained by phone call at D28 (if the patient has been discharged before D28) and at D90.

Inclusion Criteria:

• Age ≥ 18 years old
• COPD (according to GOLD 2020), whatever the stage (I-IV)
• Acute exacerbation (defined as the onset or worsening of one or more of the usual signs/symptoms of COPD) with acute worsening of respiratory symptoms that result in additional therapy) with acute respiratory failure requiring admission to ICU and ventilatory support (invasive mechanical ventilation or non-invasive mechanical ventilation or high-flow nasal oxygen therapy with FiO2 ≥ 50%)
• Informed consent of patient, patient's immediate family/ or inclusion in an emergency situation
• Affiliation to a social security

Exclusion Criteria:

• The interval between admission to the hospital and admission to ICU more than 3 days
• Antibiotic therapy clearly needed for a suspected or documented extra-respiratory infection
• Congenital or acquired immunosuppression (congenital immune deficiency, high-grade hematologic malignancies, use of immunosuppressive drugs in the last 30 days including anti-cancer chemotherapy and antirejection medications, corticosteroid treatment ≥ 20 mg/d prednisone equivalent for at least 14 days, neutropenia, HIV with unknown or known CD4 <200 / µL in the past 6 months)
• Tracheotomy
• Bronchiectasis / cystic fibrosis
• Moribund patient (imminent death)
• Patient deprived of liberty and / or under legal protection measure
• Patient already included in MULTI-EXA
• Patient already included in a type 1 interventional study on antibiotics
• Ongoing pregnancy

De-identified individual-participant data will be made available outside the primary research group for secondary research purposes like re-analysis, secondary analysis, or meta-analysis, and shared via an online secured platform.