ANTIBIOTICS EFFICACY FOR OTITIS MEDIA EFFUSION SUPPORTED BY META-ANALYSIS
Executive Summary
ANTIBIOTICS EFFICACY FOR OTITIS MEDIA EFFUSION SUPPORTED BY META-ANALYSIS, according to a report published in the April issue of Otolaryngology-Head and Neck Surgery. Summarizing a meta- analysis of 10 published studies of antimicrobial treatment of ear fluid, the report concludes that "antimicrobials are efficacious for the short-term cure of [otitis media effusion]; the effect is statistically significant, minimally affected by quality variations among studies and unlikely to represent publication bias." The report addresses the persistent debate over the meaning of published studies on antibiotics and otitis media with effusion. In the Dec. 18 Journal of the American Medical Association, for example, Erdem Cantekin, PhD, published a dissenting report from a trial in the mid-1980's. The Cantekin paper reported that amoxicillin is not an effective treatment for persistent asymptomatic ear effusion (fluid). The issue also carried a comment by Deputy Editor Drummond Rennie, MD, urging that meta- analysis of otitis media trials should include studies with negative efficacy results ("The Pink Sheet" Dec. 23, 1991, p. 17). Written by Richard Rosenfeld, MD, George Washington University, and J. Christopher Post, MD, Children's Hospital of Pittsburgh, the new report notes that "although 30%-50% of chronic effusions contain bacteria, the medical literature offers conflicting evidence regarding the efficacy of orally administered antibiotics" in speeding "resolution of the fluid." The study was "supported in part by a grant" from NIH, the report notes. Of 10 randomized, controlled trials reported in the literature, five report significant benefit and five report no significant benefit over placebo or no treatment. The 10 studies were conducted by 10 different investigative teams testing six antibiotics to treat 1,325 children in five countries. An additional seven studies, "which uniformly supported the use of antibiotics" to treat otitis media effusion, were excluded from the meta-analysis. Reasons for exclusion were "absence of concurrent controls, failure to randomize, use of a prophylactic antibiotic, simultaneous administration of an oral steroid, bilateral pretreatment tympanocentesis and comparison with controls who received an antihistamine-decongestant preparation." Rosenfeld and Post suggest that pending new evidence showing a relationship between antimicrobial agent and therapeutic outcome, treatment should begin with the least expensive antibiotics -- amoxicillin (SmithKline Beecham's Amoxil) or, for children allergic to penicillin, trimethaprim with sulfamethoxazole (TMP/SMZ -- Roche's Bactrim or Burroughs Wellcome's Septra). On the other hand, the authors did not recommend erythromycin as a single-agent therapy for otitis media because of its "variable activity against H. influenzae." They added that erythromycin was "the only drug for which the control group responded more favorably." "Results favored the treatment group in all trials except Moller et al.," the article states, noting that the Moller trial was "the only study to use erythromycin alone." The report notes that "the remaining trials showed considerable overlap in treatment effect" and therefore suggested that "no single antimicrobial regimen was superior." Antibiotic therapy benefited slightly more than one in five patients, the study found. "Children treated with antibiotics cleared their effusions on average 22.8% more often...than those who received placebo or no drug," the authors said. They reported "a statistically significant difference between outcomes in the treatment and control groups (p < 0.05)." Although the 22% rate is not dramatic, the article indicates that antibiotic therapy should precede more invasive procedures to treat effusion. The authors found that "the effect of antimicrobials on otitis media effusion is clinically and statistically significant." Children treated for 10-30 days "were between two and four times more likely to experience complete resolution [of effusion] than controls who received either placebo or no drug," they said. "The increase in cure rate achieved by antibiotics for all trials combined may be estimated as 22.0% after outliers are excluded." They estimated that "approximately 50% of the cures achieved in the pooled sample of 1,325 children were attributable to antimicrobial therapy."