How should zoster trials be conducted?

In 1994, an international group of interested clinicians and biostatisticians met to discuss the design of clinical trials in herpes zoster. They agreed that trials in herpes zoster should have prospectively agreed definitions of all outcome measures and plans for data analysis. In immunocompetent individuals, in whom pain is the major outcome measure, trials should only include patients over the age of 50 years, and for those recruited within 72 h of rash onset, should be designed to demonstrate superiority of any new therapy over existing antivirals. The primary endpoint should be time to cessation of pain for at least 4 weeks and, for the purposes of statistical analysis of its duration, the pain associated with herpes zoster ought to be considered as a continuum. All other variables, including the incidence of post-herpetic neuralgia and effects upon quality of life should be considered as secondary end-points. Evaluation of treatment effects on primary endpoints should be based upon an intent-to-treat (ITT) analysis and subgroup analysis should be used only to support the findings of the ITT analysis. These elements of good study design should be borne in mind in the evaluation of current and future trails of antiviral drugs in herpes zoster.