ABSTRACT
Secondary peritonitis includes community-acquired and nosocomial peritonitis. These intra-abdominal infections have a common pathogenesis but some microbiological differences, particularly with respect to the type of bacteria recovered and the level of antimicrobial susceptibility. This report describes a prospective observational study of 93 consecutive patients with secondary peritonitis during an 11-month period. Community-acquired peritonitis accounted for 44 cases and nosocomial peritonitis for 49 cases (post-operative in 35 cases). Fifteen multidrug-resistant (MDR) bacteria were recovered from 14 patients. In univariate analysis, the presence of MDR bacteria was associated significantly with pre-operative and total hospital lengths of stay, previous use of antimicrobial therapy, and post-operative antimicrobial therapy duration and modifications. A 5-day cut-off in length of hospital stay had the best specificity (58%) and sensitivity (93%) for predicting whether MDR bacteria were present. In multivariate analysis, only a composite variable associating pre-operative hospital length of stay and previous use of antimicrobial therapy was a significant independent risk-factor for infection with MDR bacteria. In conclusion, knowledge of these two factors may provide a more rational basis for selecting initial antimicrobial therapy for patients with secondary peritonitis.
Subject(s)
Anti-Bacterial Agents/pharmacology , Drug Resistance, Multiple, Bacterial , Peritonitis/drug therapy , Peritonitis/microbiology , Adult , Aged , Anti-Bacterial Agents/therapeutic use , Community-Acquired Infections/drug therapy , Community-Acquired Infections/microbiology , Cross Infection/drug therapy , Cross Infection/microbiology , Humans , Incidence , Middle Aged , Prospective Studies , Risk FactorsABSTRACT
The circumstances of diagnosis of human pasteurellosis are reviewed. The diagnosis is usually suspected for animal bite or scratch wounds. Conversely, in other infections the diagnosis is only based on bacteriological data. Phenotypic misidentification of Pasteurellaceae from clinical material is common. The phenotypic criteria of identification of the six species of human pathogen Pasteurella are presented. We emphasise that bite wound specimens have to be cultured for aerobic and anaerobic bacteria and yield an average of 5 bacterial isolates per culture. Antibiotic therapy relies upon amino-penicillins or cephalosporins, although b-lactamase producing strains are scarce. Fluoroquinolones can be an alternative for systemic infections. Molecular typing unequivocally points out the risk of transmission from pets to humans. Immunocompromised persons have to be made aware of precautions.
