To be eligible for inclusion, studies had to be prospective, randomised, double-blinded, controlled, compare any treatment for Herpes simplex virus keratitis with any other treatment or placebo, use specific clinical criteria for diagnosis and outcomes and include a description of the statistical analysis used. The reported outcomes were healing defined by fluorescein or rose bengal staining, slit lamp evaluation and in one study, polymerase chain reaction for Herpes simplex virus DNA and viral culture.
In the trials of Herpes simplex virus epithelial keratitis, 16 used acyclovir ophthalmic ointment (3%) or 400mg orally, seven used trifluridine (1% to 2%) in various preparations, adenine arabinoside (five trials, 1% to 3%), bromovinyl deoxyuridine (two trials, 0.1% to 1%), epidermal growth factor (one trial, 10µg/mL), human leukocyte interferon (six trials, 107 IU/ml), idoxuridine (three trials, 0.5% to 5%), vidarabine (six trials, 3%), foscarnet (one trial, 3%), thermocautery (one trial), minimal wiping debridement (one trial) and placebo alone or in combination (11 trials).
In the trials of Herpes simplex virus stromal keratitis, acyclovir was used in six trials (3% or 400mg orally), betamethasone (two trials, 0.01% to 0.1%), trifluridine (three trials, 1%), prednisone (two trials, 1%), dexamethasone (one trial, 0.01%), flurbiprofen (one trial, 0.03%). Placebo was used in all trials, alone or in combination.
Three reviewers independently determined the eligibility of each article for inclusion. Only studies uniformly agreed upon as meeting the criteria were included.