Am Fam Physician. 2005;71(6):1208-1210
Penicillin has long been the favored antibiotic for the treatment of pharyngitis caused by group A streptococci, but some recent studies have raised concerns about possibly increasing failure rates with penicillin therapy. Two previously published meta-analyses comparing cephalosporins with penicillin suggest that cephalosporins are more likely to achieve eradication of streptococcal pharyngitis infection. Casey and Pichichero present another meta-analysis of cephalosporins and penicillin; this analysis included adult patients with strep throat.
The authors employed rigorous statistical standards to decide which studies merited inclusion in the meta-analysis to avoid some of the criticisms leveled at prior meta-analyses. Of the 66 published reports initially reviewed, only eight (including nine trials) were included in the final data analysis. All of the included studies had been performed within the past 25 years, and a variety of first-, second-, and third-generation cephalosporins were used among the different trials. A total of 2,113 adult patients from the nine trials were pooled for data analysis.
The primary outcome measured was the rate of bacteriologic cure, i.e., a negative streptococcal culture after antibiotic therapy. To sum up all the trial results, bacteriologic cure was about twice as likely (odds ratio, 1.83) with a cephalosporin than with penicillin. Point estimates in seven of the studies favored cephalosporins; four of these showed a statistically significant benefit. Two studies had a nonsignificant trend favoring penicillin.
Clinical cure, i.e., resolution of pharyngitis symptoms, also was about twice as likely with cephalosporins (odds ratio, 2.29) than with penicillin. When the authors restricted the meta-analysis to trials that passed higher quality standards (for example, double blinding and elimination of chronic streptococci carriers or test-of-cure throat cultures), the odds ratios were somewhat lower but still favored cephalosporins.
The authors conclude that bacteriologic and clinical cure of streptococcal pharyngitis is about twice as likely to occur with cephalosporin therapy than with penicillin.
editor’s note: Although the authors have gone to considerable lengths to provide a “clean” and reliable meta-analysis, their published results are missing vital data. Reporting only odds ratios provides a relative comparison of cephalosporins and penicillin, but the absolute numbers relating to cure rates and other outcome measures also are needed so that the clinical significance of any differences can be assessed. For example, if the failure rate with a cephalosporin was 2 percent while the failure rate of penicillin was 4 percent, it could be said that treatment failure was twice as common with penicillin, and perhaps it could then be inferred that penicillin should not be used. If the absolute numbers are reviewed, however, the clinical significance of a 96 percent or a 98 percent success rate would be much more debatable.
In the meta-analysis reviewed here, some individual trials favored cephalosporins while others favored penicillin, leading one to suspect that, overall, these antibiotics had largely similar effects. An accompanying editorial1 questions the importance of the 5.4 percent absolute difference in eradication rates from the pooled studies, concluding, for this and other reasons, that “penicillin at this time remains the drug of choice.”—b.z.