et al., 1988; Shapiro et al., 1988). This observation was not easily explained by differences in study design or vaccine potency or by genetic differences among study populations. Furthermore, in several case-control studies there appeared to be an increased incidence of disease in the immediate postimmunization period (7 days or less) (Black et al., 1988; Harrison et al., 1988; Osterholm et al., 1988; Shapiro et al., 1988).
PRP vaccine contained only purified capsular polysaccharide, so the early-onset cases of Hib infection could not have been caused by infectious material in the vaccine itself. Several investigators postulated that the apparent increased susceptibility to infection in the immediate postimmunization period might be related to a transient decrease in preexisting antibody caused by the formation of complexes of antigen with antibody or by transient suppression of antibody synthesis (Marchant et al., 1989; Sood and Daum, 1990). Black and coworkers (1988) postulated that the clustering of early cases in the first week after immunization and the absence of cases in the second and third weeks after immunization observed in their study suggest that immunization may shorten the incubation period of Hib in children already destined to become ill. This mechanism implies that there is a redistribution of cases of disease to earlier in the time postimmunization rather than an actual increase in the rate of disease.
Observations of decreased immunity in the immediate postimmunization period were recorded as early as 1893 by Brieger and Ehrlich in a study of immunity in goat's milk after immunization with tetanus cultures. In 1896, Salomonsen and Madsen noted a decreased anti-diphtheria antibody content in the serum of horses early after immunization with diphtheria toxin, and those workers pointed out three phases of antitoxin response: (1) a fall, later called the "negative" phase, (2) a rise, and, after phase 2 reaches a maximum, (3) a fall. Madsen and coworkers (1937) studied the "negative" phase of the antitoxin curve in detail using rabbits immunized with diphtheria toxin. They were able to demonstrate that the negative phase in their experiments resulted from fixation between antigen and antitoxin in the bloodstream. Studies of individuals immunized for typhoid fever have suggested that immunization may transiently enhance susceptibility to infection both in epidemics of typhoid fever and in experimental murine infections (Raettig, 1959; Topley, 1938).
Using an infant rat model of Hib infection, Sood and colleagues (Sood and Daum, 1990; Sood et al., 1988) showed that the passive protection of animals provided by human immunoglobulin containing 5 µg of anti-PRP