Unexpected individual clinical site variation in eradication rates of group a streptococci by penicillin in multisite clinical trials

Pediatr Infect Dis J. 2007 Dec;26(12):1110-6. doi: 10.1097/INF.0b013e31814615ac.

Abstract

Background: Previously, we reported an unexpectedly large percentage of failures by penicillin to eradicate group A streptococci (GAS) from the upper respiratory tract. Because penicillin has been the recommended therapy for the treatment of GAS pharyngitis, our report prompted controversy. Data from clinical trials in which our laboratory has participated demonstrated marked variation in GAS eradication rates among clinical sites. The reasons for such variation have never been adequately examined. We performed statistical analyses of site variation in eradication rates to assess the potential effect on reported reduced penicillin efficacy.

Methods: Penicillin GAS eradication rates were compared using data from 4 large multisite pharyngitis treatment trials (75 clinical sites; 1158 subjects). Variation in eradication rates among clinical sites was statistically evaluated [chi(2) tests and generalized estimating equation (GEE) regression models].

Results: There was significant site-to-site variation in GAS eradication rates in each of the trials (range, 17-100%; P < 0.005) as well as between separate trials (mean range, 58-69%; P < 0.033). GEE modeling indicated that GAS eradication rates were significantly higher for clinical sites participating in more than one clinical trial.

Conclusions: The statistically significant site-to-site variation in penicillin eradication rates was related to factors (dependencies) at individual sites. Such factors may affect assessment of therapeutic efficacy and indicate a necessity for considering clinical site variation before reporting pooled efficacy data from multiple sites; combined data may result in misleading clinical implications. This is the first report documenting significant variation resulting from individual clinical site-related factors and offers a possible explanation for reduced penicillin eradication.

MeSH terms

  • Anti-Bacterial Agents / administration & dosage
  • Anti-Bacterial Agents / therapeutic use*
  • Clinical Trials as Topic / statistics & numerical data*
  • Humans
  • Multicenter Studies as Topic / statistics & numerical data*
  • Penicillin G Benzathine / administration & dosage
  • Penicillin G Benzathine / therapeutic use*
  • Penicillin V / administration & dosage
  • Penicillin V / therapeutic use*
  • Pharyngitis / drug therapy*
  • Pharyngitis / microbiology
  • Randomized Controlled Trials as Topic
  • Regression Analysis
  • Research Design
  • Streptococcal Infections / drug therapy
  • Streptococcal Infections / microbiology
  • Streptococcus pyogenes / classification
  • Streptococcus pyogenes / drug effects*
  • Streptococcus pyogenes / isolation & purification
  • Treatment Failure
  • Treatment Outcome

Substances

  • Anti-Bacterial Agents
  • Penicillin G Benzathine
  • Penicillin V