Evidence for the effectiveness of topical ocular antibiotics in the treatment of infective conjunctivitis comes from randomized controlled trials in secondary and primary care.

• A Cochrane review investigated the effect of antibiotic treatment on the management of acute bacterial conjunctivitis. Central, Medline, Embase, Sigle, NRR, PubMed were searched to the start of 2006, and hand searches of bibliographies were conducted. Pharmaceutical manufacturers were also contacted as necessary [Sheikh and Hurwitz, 2006].
• Five trials (n = 1034) were found, three secondary care based (n = 527), and the two more recent trials were community based (n = 507).
• Meta-analysis of data on clinical and biological remission rates reveal that topical antibiotic use:
• In early days (2 to 5) are of benefit in improving clinical (RR 1.24, 95% CI 1.05 to 1.45) and microbiological (RR 1.77, 95% CI 1.23 to 2.54) remission rates.
• In later days (6 to 10), early advantages in clinical (RR 1.11, 95% CI 1.02 to 1.21) and microbiological cure rates are reduced (RR 1.56, 95% CI 1.17 to 2.09), but persist.
• Note: in the two primary care trials no statistically significant difference in clinical recovery from infective conjunctivitis was seen in people receiving a topical ocular antibiotic compared with placebo [Rietveld et al, 2005; Rose et al, 2005].
• Most cases however resolve spontaneously with clinical remission being achieved in 65% (95% CI 59 to 70) by days 2 to 5 in those receiving placebo.
• The authors concluded that although the use of antibiotics is associated with significantly improved rates of clinical and microbiological remission, the benefits are marginal as in most cases the infection is self-limiting. Therefore, when people with clinically diagnosed infective conjunctivitis are treated, the chances that an individual will benefit from treatment is very small.
• Of the secondary care based trials, in two trials data from people who were treated and subsequently found not to have bacterial conjunctivitis were excluded from the analysis. In the third trial treatment was delayed until microbiological results were available to guide treatment. Therefore the clinical usefulness of this information is limited because in practice it is not possible to prove that a person has a bacterial infection before starting treatment, unless treatment is delayed until microbiological results are available.