Two notes concerning the data presented in the last two slides.

(A) You will notice that the figures represent an average of the data identified over a 25 year period. This time frame is used because disasters are relatively rare events. They fluctuate in number from year to year. A long time frame helps to elimi nate misinterpretation of the data due to these fluctuations.

(B) The data most often represent those reported to relief agencies. In general, data on death and, more specifically, injuries from disasters have been called into question regarding their quality. Information on mortality and morbidity from a disaste r may be captured in existing surveillance systems or special systems established at the time of the disaster. Questions about the accuracy of the figures arise for several reasons. For example, populations are often not fully enumerated. If large numb ers of persons are missing, then one must estimate the impact from existing population records. Another example is the inability of many medical records systems to distinguish disaster-related events from other, non-disaster, events.