ABSTRACT
A major stumbling block in studying past episodes of many diseases, including influenza, is the lack of evidence on morbidity. Mortality data have their limitations but they are usually collected reasonably systematically, at least for the last couple of centuries. This is not the case for morbidity data, and this is particularly true of influenza. Where morbidity data are found, they are often simple and have a number of limitations. First, they are likely to be an under-estimation and, second, they usually have no temporal or spatial element – there is no detailed intelligence regarding where or when the cases occurred. Influenza is often self-diagnosed and self-treated, and only a small proportion of cases ever come into contact with the medical professions, and even then diagnosis is not certain. These problems are as true now as ever. Consequently, modern surveillance of influenza is often based on serological testing and pathology of a sample of the population. Little information relating to morbidity during the 1918-19 pandemic exists, and when it does it tends to be anecdotal.
