Azithromycin compared with beta-lactam antibiotic treatment failures in pneumococcal infections of children

Abstract

Objective: To determine whether treatment failures occurred more commonly with azithromycin than with beta-lactam antibiotics in children who developed invasive pneumococcal disease within 30 days of receiving prior antimicrobial therapy.

Methods: Retrospective review of medical records of children evaluated at Texas Children's Hospital between 1996 and 2002 who had received antimicrobials (azithromycin or a beta-lactam antibiotic) and developed invasive pneumococcal disease within 30 days. Treatment failure was defined as invasive pneumococcal infection that occurred while taking antimicrobials or within 3 days of stopping azithromycin treatment or 1 day of stopping beta-lactam treatment. Penicillin and azithromycin susceptibilities were determined and categorized according to National Committee for Clinical Laboratory Standards guidelines.

Results: We identified 21 and 33 children with similar demographic features who had developed invasive pneumococcal disease within 1 month of receiving azithromycin or a beta-lactam antibiotic, respectively. Eleven (52%) children in the azithromycin group and 11 (33%) in the beta-lactam group met the definition for treatment failures (P = 0.34). Eight treatment failures while receiving azithromycin were caused by pneumococci with the macrolide-resistant (M) phenotype, 2 with the macrolide-, lincosamide- and streptogramin B-resistant (MLSB) phenotype and 1 by a macrolide-susceptible organism. In the beta-lactam group 7 had a penicillin-resistant isolate, 3 had an intermediately susceptible isolate and 1 had a susceptible isolate.

Conclusions: Our study suggests that treatment failures among patients who developed invasive disease within 30 days of receiving an antimicrobial occur as frequently in patients who receive beta-lactam antibiotics as in those who receive azithromycin. Furthermore macrolide resistant organisms are not more likely to be recovered after a macrolide treatment failure than a penicillin-nonsusceptible isolate being recovered after a beta-lactam treatment failure (P = 1.0).