Variola minor in Bragança Paulista county, 1956. Attack rates in various population units of the two schools including most students with the disease.

Overall attack rates were different for the two schools but were the same when only classes with cases were considered. Of the 49 classes (from both schools) with susceptible students, 29 had cases. In these, the attack rate among susceptibles ranged from 4-69%. The differences between the class rates were highly statistically significant. In the José Guilherme School, 17 out of 21 classes containing susceptibles had cases as compared with 12 out of 28 classes in the Jorge Tibiriçá School. Attack rates differed little between schools for either the early or late shift classes, but the attack rate for early shift classes in both schools together (39%) significantly exceeded that for the late shifts (16%). Within-shift attack rate differences were highly significant for the first-shift classes but not significant for classes in the other shifts. Attack rates differed significantly by classroom population (group of students using a classroom including all shifts), mainly because of heterogeneous rates in one school. Attack rates by desk column did not differ significantly even when comparing the column seating the earliest, the latest or the "previous" case in the class with the remaining columns. Attack rates by desk row also did not differ significantly. There was no important progressive association between attack rates and size of "unit of susceptibles" (group combining columns and rows), but there was a significant difference for unit size. Attack rates were not influenced by cluster seating of susceptibles.