[Potency testing of inactivated rabies vaccines. Problems posed by the use of the Habel test]

Rabies vaccines containing inactivated virus are often tested for potency by the Habel test. Even though many people use the Habel test, a humpbacked curve of mortality is always observed for the vaccinated mice, with a linear regression for the control animals. This unexpected survival of vaccinated mice given the greatest doses of virus is an artefact since the phenomenon is not observed after a non intracerebral challenge. Since the intracerebral challenge is not a good model for natural infection, the index of protection for the tested vaccines must be evaluated after discarding aberrant values by a graphic method. The use of the Reed and Muench method or Kärber method leads to an overestimated index of protection. The N.I.H. test gives the best estimation of the effectiveness of rabies vaccines, but does not reproduce the natural conditions of vaccination and infection.