ABSTRACT
This chapter discusses how data for surveillance are collected and interpreted, and identifies several sources of potential bias. The notification systems were not instituted as a purely epidemiological exercise, but rather to make possible the rapid application of preventive measures. The main task of a surveillance system is to discover sudden changes in incidence, that is, outbreaks or epidemics. Surveillance is an important part of practical infectious disease epidemiology. It builds on the notion that whilst any physician may only see one or two cases of an outbreak, and thus not be aware of it, the collected notifications on a regional or national level will make it possible to see the whole picture. Attendance patterns, diagnostic methods, screening, and reporting propensity are all examples of biases that could enter notification data, and they underline the caution with which such data should be approached.
