Published online by Cambridge University Press: 15 May 2009
Cases of paratyphoid fever are often not diagnosed until the second, third or subsequent weeks of illness. When calculating clearance rates of a series of cases the calculations must be based only on the numbers known to be positive at the week under consideration. If based throughout on the total number of cases the rates of clearance in the early weeks are greatly reduced.
Cases for which laboratory records are incomplete must not be entirely rejected when calculating clearance rates but must be retained in the population for as long as they were known to be positive.
Analyses designed to show the duration of infection in paratyphoid fever can only be made with accuracy under the most favourable conditions.
A large number of cases of paratyphoid fever were repeatedly examined bacteriologically to establish the duration of the infection as distinct from the clinical illness. After an initial lag the proportion of cases remaining infected fell logarithmically until the carriers revealed themselves.
We are grateful to many medical officers for information about their cases, and to Dr Lewis Fanning and Dr Ian Sutherland for much helpful criticism.