Aside from the interpretation of the results there is possibly a more serious reason for objecting to the beginning of vaccination during an epidemic. This lies in the danger of producing a temporary negative phase in the patient, which makes him somewhat more susceptible to natural infection for a few hours immediately following each administration.
McCoy (15) outlined the requirements necessary for an ideal vaccine experiment as follows: 1. The community should be as large as possible, and should number at least 10,000 persons. 2. The conditions under which they live should be as nearly equal as possible. 3. The turnover, or rather the change in population, should be as small as possible. 4. The social service should be efficient and reliable, so that it can be definitely ascertained when anyone becomes sick and what the disease is from which he is suffering. 5. Fifty per cent. should be vaccinated before the epidemic arrives, and the other 50 per cent. should be held as controls.
No examples were found which came up to the above requirements, but there were some instances in which vaccination was completed before the epidemic appeared, and some in which we were able to get a week-by-week comparison between vaccinated and unvaccinated groups. Most of the data which has been reported shows that vaccination was begun about the last of the second or the first of the third week of the epidemic, and in some instances not until after the peak was passed. Add to this the fact that the vaccine was given in from three to four doses, at from three to seven day intervals—a course which required in the neighborhood of two weeks for completion—and it is obvious that the full protective powers of the vaccine were not acquired by the individual until the worst of the epidemic was over and the number of cases were rapidly subsiding.
In order to get the best understanding from these experiments, the data will be presented in three series: I. Those instances in which vaccination was completed before the epidemic appeared. II. Those instances in which it is possible to compare the relative occurrence of influenza in both the vaccinated and unvaccinated groups after vaccination was completed. III. Those instances in which vaccination was begun after the epidemic appeared and in which comparisons of total figures only are available.