The study was stopped for recruitment defect

The study aims to evaluate a deescalating therapeutic strategy (switch the carbapenem to another beta-lactam for which the isolated pathogen is susceptible) in patients with well-defined ESBL-PE infections (usual sites of infections and non severe infections).

Carbapenems are considered the antibiotics of choice for the treatment of infections due to expanded-spectrum beta-lactamase producing Enterobacteriaceae (ESBL-PE). Their use is readily increasing because of the pandemic of ESBL-PE. However, the future of these drugs is challenged by the worldwide emergency of new resistance mechanisms in the same pathogens, particularly the carbapenemases. This resistance makes these drugs inactive and gives no therapeutic options for patients. Resistance to carbapenems is strongly associated with the use of these drugs, because of their large spectrum and strong impact on the host flora. Consequently, the correct use of these drugs is an important public health objective. In France, the CA-SFM (Committee on Antimicrobial of the French Society for Microbiology) has suggested alternative strategies with narrow spectrum beta-lactam for the treatment of ESBL-PE infections, including 3rd or 4th cephalosporins, cefoxitin and penicillins with beta-lactamase inhibitors. A recent analysis including 6 prospective cohort studies of bloodstream infections due to ESBL-producing Escherichia coli did not find an excess of mortality or a longer length of hospital stay for patients receiving penicillin with beta-lactamase inhibitors, as compared to patients treated with carbapenems, after adjusting by confounding factors. As patients receiving carbapenems seem to be more severely affected, the efficacy and safety of the alternative strategy remain unproved. A randomized study is therefore needed to evaluate a deescalating therapeutic strategy (switch the carbapenem to another beta-lactam for which the isolated pathogen is susceptible) in patients with well-defined ESBL-PE infections (usual sites of infections and non severe infections). The objectives of the present study are to demonstrate that the alternative strategy of deescalating therapy is not inferior to a strategy maintaining the carbapenem in terms of clinical cure, survival, lack of relapse and microbiological cure.

Inclusion Criteria:

• Age ≥ 18 years
• Hospitalization in conventional ward; (3) Receiving a curative treatment with a carbapenem for at least 24 hours and less than 3 days for a recent infection due to an ESBL-PE, either initially or secondary documented
• With a site of infection originating from the urinary, digestive or biliary tract
• Identification of an ESBL-PE for which susceptibility results (by the method of discs) have shown to be susceptible to more narrow spectrum beta-lactams (cephalosporins, b-lactamase inhibitors, monobactams)
• With sepsis signs and symptoms controlled after initiation of antibiotic therapy
• For a community-acquired or hospital-acquired infection.

Exclusion Criteria:

• Pregnancy or breastfeeding
• Neutropenia (PNN < 500/mm3)
• Hospitalization in intensive care unit or bone marrow transplant unit
• Documented polymicrobial infection
• Culture of an ESBL-PE susceptible to an orally active drug (such as fluoroquinolones, cotrimoxazole) and possible use of one of these drugs
• Need to treat a EBLSE infection(s) wtih more than drug other than carbapenems
• Need to maintain an association with an aminoglycoside
• Colonization without signs and symptoms of sepsis
• Sepsis signs and symptoms not controlled at the time of enrolment
• Known allergy to beta-lactams
• Failure to complete medical examination
• Absence of signed written consent.
• Patient without healthcare insurance (French social security, CMU or AME)