Inappropriate antibiotic therapy in ventilator-associated pneumonia (VAP) is associated with increased mortality. The international guidelines recommend using broad spectrum antimicrobials especially in patients who received previous antimicrobials, with risk factors of muti-drug resistant (MDR) VAP or after 5 days of mechanical ventilation. Using broad-spectrum antibiotics for 48h until the results of conventional cultures and antimicrobial susceptibility testing (AST) are available, may promote the emergence of drug-resistant bacteria. Exposure to imipenem, as short as 1 to 3 days, is associated with a 5-fold increase in the risk of imipenem resistance in the gut microbiota of ICU patients (Armand-Lefevre AAC 2013). Performing AST directly on clinical respiratory samples would hasten the process by at least 24h.

The diagnostic performance of a rapid method combining mass spectrometry and direct AST [DAST] are previously analyzed, and compared it with the conventional method (mass spectrometry with conventional AST [CAST]) and its potential impact was assessed on antimicrobial use in 85 patients (Le DORZE M et al - Clin. Microbiol. Infect. 2015).

The results produced by the dast were useable in 85,9% of the cases and the sensitivity and negative predictive values of DAST were 100% for all antibiotics tested, except gentamicin (97.1% [95%CI = 93.3-101] and 97.4% [93.7-101], respectively) and amikacin (88.9% [81.7-96.1] and 96.4% [92.1-100.7], respectively), compared with CAST. Specificity and positive predictive values ranged from 82.9 (74.2-91.5) to 100%, and from 86.4 (78.5-94.2) to 100%, respectively. If results had been reported to the clinicians, that DAST would have saved carbapenem prescription in 17 cases (22%) and would have allowed immediate narrow spectrum antimicrobials in 35/85 (41.2%) cases. But, the benefit of DAST was based on a simulation and should be now tested in a randomized fashion. This project is a prospective multicenter study. The hypothesis is that, DAST compared to CAST, would increase the number of adequate antimicrobial therapy within 24 hours in case of late VAP (> 5 days under mechanical ventilation) with Gram negative bacilli (GNB) in IC patients while sparing carbapenems (imipenem and meropenem). The primary objective is to determine the impact of a strategy using DAST on the rate of day1 adequate therapy without carbapenems in case of late VAP due to GNB.

Inappropriate antibiotic therapy in ventilator-associated pneumonia (VAP) is associated with increased mortality. The international guidelines recommend using broad spectrum antimicrobials especially in patients who received previous antimicrobials, with risk factors of muti-drug resistant (MDR) VAP or after 5 days of mechanical ventilation. Using broad-spectrum antibiotics for 48h until the results of conventional cultures and antimicrobial susceptibility testing (AST) are available, may promote the emergence of drug-resistant bacteria. Exposure to imipenem, as short as 1 to 3 days, is associated with a 5-fold increase in the risk of imipenem resistance in the gut microbiota of ICU patients (Armand-Lefevre AAC 2013). Performing AST directly on clinical respiratory samples would hasten the process by at least 24h.

The diagnostic performance of a rapid method combining mass spectrometry and direct AST [DAST], are previously analyzed and compared it with the conventional method (mass spectrometry with conventional AST [CAST]) and its potential impact was assessed on antimicrobial use in 85 patients (Le DORZE M et al - ICAAC 2013, and clinical microbiology and infections submitted).

The results produced by the dast were useable in 85,9% of the cases and the sensitivity and negative predictive values of DAST were 100% for all antibiotics tested, except gentamicin (97.1% [95%CI = 93.3-101] and 97.4% [93.7-101], respectively) and amikacin (88.9% [81.7-96.1] and 96.4% [92.1-100.7], respectively), compared with CAST. Specificity and positive predictive values ranged from 82.9 (74.2-91.5) to 100%, and from 86.4 (78.5-94.2) to 100%, respectively. If results had been reported to the clinicians, that DAST would have saved carbapenem prescription in 17 cases (22%) and would have allowed immediate narrow spectrum antimicrobials in 35/85 (41.2%) cases. But, the benefit of DAST was based on a simulation and should be now tested in a randomized fashion.

Hypothesis - DAST as compared to CAST, would increase the number of adequate antimicrobial therapy within 24 hours in case of late VAP (> 5 days under mechanical ventilation) with Gram negative bacilli (GNB) in ICUs while sparing carbapenems (imipenem and meropenem).

Primary objective - To determine the impact of a strategy using DAST on the rate of day1 adequate therapy while sparing carbapenems in case of late VAP due to GNB.

Secondary objectives are:

1. To evaluate the impact of the DAST strategy compared to CAST on the following outcomes:

   • Adequate antimicrobial therapy at Day1 (with or without carbapenem)
   • exposure to carbapenems at D14 and D28
   • Broad spectrum antibiotic exposure at D14 and D28
   • De-escalation at D1
   • relapses or new VAP at D14 and D28.
   • the emergence of ESBL producing GNB and / or resistant to broad-spectrum antibiotics in clinical samples and in the intestinal microbiota at D3, D7, D14 and D28, or until discharge of the ICU patient
   • Changes in respiratory infection and organ failure
   • Mechanical ventilation exposure between day 1 and 28
   • Length of ICU stay.
   • In-ICU mortality.

2. To evaluate the performance diagnostic of DAST compared to CAST

3. To achieve a minimization study of the costs of the DAST strategy compared to the CAST strategy Experimental scheme and procedures This study is a multicentre, prospective, parallel-group, open-label trial. After baseline screening, an independent, centralized, computer-generated randomization sequence will be used to randomly assign patients in a 1:1 ratio to DAST or CAST groups. Random allocation will be stratified by centre; investigators will be masked to assignment before, but not after, randomization, as per our open-label design. This system will be password protected and access by the principal investigator or study coordinator after the patient or surrogate give consent and has met inclusion criteria. Although treatment assignments were not masked, all investigators were unaware of aggregate outcomes during the study.

Number of patients According to our preliminary study, 76 patients could be enrolled within one year in 3 ICUs and more than half will receive carbapenems initially. DAST would have permit to spare carbapenems in 22% of the cases and to avoid inadequate penem use in 5 other percent of the cases. The expected proportion of adequate antibiotic therapy without carbapenem is 68%. In all, 68 patients per arm will be necessary to detect a 22% difference between groups with a power of 90% (pooled variance Z test). The level of significance will be 5%.

The hypothesis is that 15% of DAST will be non-contributive and if 5% patients are lost to follow up, 90 patients per arm (140 patients overall) need to be enroll.

According to previous number in the preliminary study, 180 patients will be enrolled within 18 months of inclusion.

Inclusions - 180 patients will be enrolled during a 18 months period In all, 9 investigation centers are associated to the study. Thus, around 1.1 patient need to be enrolled per month during 18 month for each center Statistical analyses - Both group will be compared by Cochrane Mantel HAENTZEL chi square for primary end point.

Other secondary objectives will require model for longitudinal data, model for censored data that account for competing events (MDR acquisition, day 28 outcome).

The random unit is the patient. The first episode will be taken into account for the primary end point.

In case of repeated VAP with GNB, samples will be proceeded according to their random allocation arms (CAST of DAST) until day 28 or ICU discharge.

Data collected and definitions:

After obtaining an appropriate respiratory specimen for Gram stain and conventional culture, initial antibiotic therapy (day 0) will be given according to the standards of care, in agreement with international guidelines: Initial Empiric Antibiotic Therapy for patients with late-Onset Disease or risk factors for MDR pathogens from ATS and IDSA 2005 (table 1-2-3) avoiding carbapenem whenever possible.

The following data will be collected :

At admission to the ICU : age, sex, pre-existing comorbidities, previous location before admission, admission category, reason for admission, SAPS II, the presence and type of organ dysfunction using the sequential organ-failure assessment (SOFA) score, and use of mechanical ventilation.

Between D1 and D3 : Microbiological results of the VAP episode (GNB identification, CAST and DAST) Between D1 and D28 All infections occuring and the colonization by MDR SOFA score and type of organ or system failure and CPIS score (D1,2,3,7,14,28) All antibiotics exposure.

Adequacy of the initial empirical antibiotics regimen for VAP suspected to be due to gram negative organisms at inclusion will be recorded.

Inadequate antimicrobial treatment was defined as the use of antibiotics to which at least one cultured isolate was resistant in vitro at day 1.

Besides the initial respiratory specimen at inclusion, all other microbiological assessments from specimens taken between day 0 = inclusion and day 14 for will be recorded. Multidrug-resistant bacteria were defined according to MAGIORAKOS Clin Microbiol infect, 2012. Broad spectrum antimicrobials were defined as antibiotics belonging to ranks 4 to 6 (Piperacillin+Tazobactam, Ticarcillin+Clavulanic Acid 4th generation cephalosporin, Antipseudomonal 3rd generation cephalosporin, Carbapenems) (Weiss et al, Clin. MICROB. Infect. 2015). For the purpose of this study, de-escalation was considered if the pivotal antibiotic was switched for an antibiotic with a narrower spectrum than the empirical treatment, according to a ranking of molecules from a recent study performed on behalf of the ""de-escalation"" study (Weiss et al, Clin. MICROB. Infect. 2015)"

Inclusion Criteria:

• Adults (18 years or older)

• Need for mechanical ventilation expected for at least 5 days at any time during the ICU stay (including if the intubation was performed before the ICU admission)

• VAP suspected and clinical respiratory samples with GNB at direct smear examination

• At least one condition with a risk factor of multidrug resistant infection:

  1. Previous use of antimicrobials (at least 2 days in the past 7 days)
  2. Risk of colonization or infection with MDR or XDR bacteria within 3 months

• Written informed consent has to be obtained from the patients or a surrogate. The patient or his surrogate can withdraw from the study at any time.

Exclusion Criteria:

• Pregnant or lactating women
• VAP suspected and respiratory samples without GNB at direct smear examination
• VAP that occurred without neither previous antimicrobial exposure in the past 5 days or neither risk of MDR colonization
• Samples send to the lab during night and weekend in center if respiratory samples are not performed during this period
• Active therapeutic limitation
• Known allergy to antibiotics
• Social welfare unavailable