Statistical certification of eradication of poliomyelitis in the Americas

Abstract

The last confirmed case of paralytic poliomyelitis due to indigenous wild poliovirus in the Americas occurred in Peru in 1991. In 1994 the International Commission on Polio Eradication of the Pan American Health Organization (PAHO) deemed eradication of polio from the area to have occurred, based on its strategic efforts and the observed results. A mathematical model is presented here which relates the time elapsed since that last detected case of paralytic poliomyelitis caused by wild poliovirus to the probability that the transmission of indigenous wild poliovirus has been stopped. The appropriateness of applying the model to various geographical areas of the Americas is investigated using data about the occurrence of confirmed cases of polio since 1984, the time of the eradication initiative adopted by PAHO. The model suggests that if four year have elapsed since the last reported confirmed case of polio caused by wild poliovirus, and no other confirmed cases have been identified, the probability of undetected indigenous wild poliovirus transmission is less than 5%. An important assumption is that the eradication strategy implemented by PAHO has yielded steady improvements. A consequence of this approach is that the annual probabilities of persistence given by the model are conservative, in the sense of being higher than the true, but unknown a priori probabilities, and more so with each passing year. It is thus seen that the model results are compatible with the conclusion reached by PAHO in 1994. The model takes into account the intensity of surveillance of each country in the region, measured by the corresponding rates of acute flaccid paralysis (AFP). Because importations of wild poliovirus may occur from other regions of the world, surveillance efforts are being maintained in the Americas until global eradication has been achieved.