The promotor conducted a multicenter prospective study, IRIS (Interest of integrons as predictive markers of acquired antibiotic Resistance in patients with urinary or Intra-abdominal Sepsis) which showed that integrons have a high negative predictive value (NPV) for antibiotic resistance. Among the 343 patients admitted in the ED or intensive care units (ICU) with a urinary sepsis, the NPV of integrons, by detecting them directly form urine samples, was 96.6% (CI:94.0-98.6%) for resistance to 3GC. Whether the detection of integrons to predict antibiotic resistance could efficiently guide empirical antibiotic therapy of patients with a UTI remains to determine. INVICTUS aims at providing the front-line ED physician with additional valuable information using molecular detection of integrons, to best guide adequate first-line antibiotic therapy, while sparing large broad-spectrum antimicrobial agents. In the experimental arm, empirical parenteral antibiotic therapy will be guided based on the detection of integrons. If integrons detection is positive, a large broad spectrum therapy with ureidopenicillin and β-lactamase inhibitors combinations (Piperacillin-Tazobactam) or carbapenems (Imipenem, Meropenem) will be prescribed. If integrons detection is negative, patient will receive a parenteral 3GC (Ceftriaxone, Cefotaxime, Ceftazidime or Cefepime). In the control arm, the comparative treatment will be an empirical antibiotic therapy with ureidopenicillin and β-lactamase inhibitors combinations (Piperacillin-Tazobactam) or carbapenems (Imipenem, Meropenem) at clinician's discretion. For the primary analysis, the number of days with large broad-spectrum antibiotic therapy during the first week will be compared between the two groups using a Student t-test.
