Context: Macrolide antibiotics, including erythromycin, clarithromycin, and azithromycin, are the mainstays of empirical pneumonia therapy. Macrolide resistance among Streptococcus pneumoniae, the most common cause of community-acquired pneumonia, is increasing in the United States. Whether resistance is a significant problem or whether macrolides remain useful for treatment of most resistant strains is unknown.
Objective: To examine the epidemiology of macrolide-resistant pneumococci in the United States.
Design and setting: Analysis of 15 481 invasive isolates from 1995 to 1999 collected by the Centers for Disease Control and Prevention's Active Bacterial Core surveillance system in 8 states.
Main outcome measures: Trends in macrolide use (1993-1999) and resistance and factors associated with resistance, including examination of 2 subtypes, the M phenotype, associated with moderate minimum inhibitory concentrations (MICs), and the MLS(B) phenotype, associated with high MICs and clindamycin resistance.
Results: From 1993 to 1999, macrolide use increased 13%; macrolide use increased 320% among children younger than 5 years. Macrolide resistance increased from 10.6% in 1995 to 20.4% in 1999. M phenotype isolates increased from 7.4% to 16.5% (P<.001), while the proportion with the MLS(B) phenotype was stable (3%-4%). The median erythromycin MIC (MIC(50)) of M phenotype isolates increased from 4 microg/mL to 8 microg/mL. In 1999, M phenotype strains were more often from children than persons 5 years or older (25.2% vs 12.6%; P<.001) and from whites than blacks (19.3% vs 11.2%; P<.001).
Conclusions: In the setting of increasing macrolide use, pneumococcal resistance has become common. Most resistant strains have MICs in the range in which treatment failures have been reported. Further study and surveillance are critical to understanding the clinical implications of our findings.