Analysis of Incubation Period and Spatial Clustering of a Common Source Outbreak

Abstract

Background and purpose: The incubation period is a characteristic common to any disease, especially infectious diseases. To ascertain a simultaneous exposure and a spatial clustering are helpful in evaluating an outbreak. Spatial distribution has played a central role in the epidemiology of a number of diseases. Facing a novel disease, for example severe acute respiratory syndrome, traditional epidemiological approaches are applicable. Methods: Sartwell's lognormal model was used to estimate the incubation period of the severe acute respiratory syndrome (SARS) outbreak on an airplane. A normal probability-probability plot of the logarithmic transformation of individual incubation periods was graphed and validated by the Kolmogorov-Smirnov test. A test of variance was applied to test the Poisson distribution, which was the description of objects distributed spatially at random. Results: Based on Sartwell's model, the estimated median of the incubation period was 37.8 hours. The normal probability-probability plot and Kolmogorov-Smirnov test established the lognormal model of the incubation period. The transmission of SARS on the airplane was also shown to be due to a common source according to this model. The Poisson distribution was rejected, indicating there was spatial clustering. Discussion: With the lack of a molecular work-up, traditional epidemiological approaches are appropriate in the investigation of an outbreak or an emerging disease.