Because tuberculosis has a variable and generally long average incubation period, changes in incidence are rarely abrupt and the multitude of discrete and usually small outbreaks are rarely becoming prominent unless the epidemic has receded to very small levels.
On can thus compare the number of cases in one period of time to the same previous period. In this example, it is the cumulative number of cases over a whole year, but it might be a shorter period of time such as weeks as it was done in the example of the 1985 US data (see next slide).
This approach of comparing the fractional change from one period compared to the previous one will also show soemthing about the consistency in the surveillance system as will shown after the example from the United States.
