UMHS LOGOUniversity of Michigan
Department of Pediatrics

Evidence-Based Pediatrics Web Site

Topical Antibiotics Are Effective in Bacterial Conjunctivitis


  • Does the use of empiric topical antibiotic therapy shorten the clinical course of acute conjunctivitis?

Clinical Bottom Lines

  1. Acute conjunctivitis is a self-limited disease, although topical antibiotic therapy with polymyxin-bacitracin shortens the duration of clinical disease and enhances eradication of the causative organism from the conjunctiva.
  2. Among patients treated with polymyxin-bacitracin, 62% showed clinical cure at 3-5 days, whereas only 28% of those treated with placebo were cured (ARR=34%, NNT=3).1  There was no difference in clinical cure rate at 8-10 days.
  3. The percentage of patients with negative cultures at 3-5 days was 71% for those treated with antibiotics and 19% for those given placebo (ARR=52%, NNT=2).  This difference was still present at 8-10 days.

Summary of Key Evidence

  1. In this randomized, double blinded, multi-center study, 116 patients were randomized to a 7 day course of either polymyxin-bacitracin ointment or placebo. At one center, patients were randomized after cultures become positive for either S. pneumoniae or H. influenzae were also treated with systemic antibiotics for associated bacterial infections. Placebo and control groups had essentially similar characteristics at the onset of the study.1
  2. Outcomes were measured during therapy (3-5 days) and at the conclusion (8-10 days) of treatment.  Outcomes included clinical cure and bacteriologic cure (i.e. eradication of the organism on culture.).
  3. No statistically significant difference in clinical cure was noted at the second visit, although no power calculation was made to determine of the sample size was large enough to detect a difference.  At day 3-5, there was a statistically significant degree of clinical cure in those receiving antibiotics.  There was an impressive difference in bacteriologic cure both at day 3-5, and at the conclusion of treatment. 
  4. Concerns regarding the validity of the data offered in the study include the following: 1) A fairly small sample size of approximately 30 patients per group. 2)  The fact that some patients were not randomized until after cultures had come back positive. 3) Analysis of the data in groups of patient different than the groups to which they were randomized. Data was analyzed in 3 groups, including one for patients who also received oral antibiotics.  This group should either have been excluded from the study, or randomized with stratification taking this variable into account.  Nevertheless, examination of the data specifically related to the treatment of bacterial conjunctivitis with topical antibiotics reveals it to most likely be significant.

Additional Comments

  • An earlier controlled study by the same author showed that 42% of cases are caused by H. influenzae, 12% by S. pneumoniae, 20% by adenoviruses, with 28% culture negative.2
  • The best clinical clues as to the presence of bacterial versus viral conjunctivitis are as follows: a) associated otitis media is suggestive of bacterial disease and warrants oral antibiotic therapy; b) associated pharyngitis is highly suggestive of adenoviral conjunctivitis (Positive Likelihood Ratio of 9); c) the presence of purulent exudate is somewhat suggestive of bacterial conjunctivitis. (Positive Likelihood Ratio of 2).2
  • Neonatal conjunctivitis involves a different spectrum of organisms (i.e. Chlamydia trachomatis and Neisseria gonorrhea) requires a culture and systemic antibiotics: the this article is not relevant to that patient population.


  1. Gigliotti F, Hendley JO, Morgan J, Michaels R, Dickens M, Lohr, J. Efficacy of topical antibiotic therapy in acute conjunctivitis in children. J of Pediatrics. 1984;104:623-6.
  2. Gigliotti F, Williams WT, Hayden FG, Hendley JO. Etiology of acute conjunctivitis in children. J of Pediatrics. 1981; 98: 531-536.

CAT Author: Timothy Cotts, MD

CAT Appraisers: John Frohna, MD

Date appraised: August 30, 1999

Last updated February 16, 2003
Department of Pediatrics and Communicable Diseases
© 1998-2002 University of Michigan Health System