Traditional phases of vaccine testing
| Phase | Objective | Outcomes | Design | Study population |
|---|---|---|---|---|
| I | Safety Often first in human administration of vaccine candidate. | Safety Immunogenicity | One or several experimental arms, ±placebo control | Healthy volunteers (<100) |
| II | Exploratory efficacy. | Immunogenicity (safety±direct protection) | Randomised controlled trial, ±placebo control (phase IIB) | Healthy volunteers or at-risk individual (hundreds, thousands) |
| III | Confirmatory Efficacy. Usually required for licensing* | Direct protection (safety, ± immunogenicity) | Placebo-controlled randomised trial Sometimes, the control group is administered an unrelated vaccine† | At-risk individuals (thousands to tens of thousands) |
| IV | Post-licensure surveillance and effectiveness, cost-effectiveness | Total protection (direct protection + herd immunity), safety (rare adverse events, long-term safety). Cost | Randomised trial (pragmatic trials), cohorts, case–control studies | Target population |
Expanded from Farrington and Miller.17
*Rarely, vaccines have been marketed based on data from phase II trials, such as in the implementation of the meningococcal serogroup C vaccine.57
†An increasingly common strategy is to offer control participants an effective vaccine that will not interfere with the immunological evaluation, thus enabling double-blind evaluation, while also providing some benefit to the control group.58–60