A model-based approach to characterize the incubation period of paediatric transfusion-associated acquired immunodeficiency syndrome

Traditional statistics, such as the average, currently provide a misleading statistical description of the incubation periods of paediatric transfusion‐associated acquired immunodeficiency syndrome (TA‐AIDS). Two types of length‐biased sampling, left censoring and right censoring, occur in the 37 cases diagnosed before 1986 and reported to the Centers for Disease Control by 15 September 1986. To correct for these problems, we propose a truncated Weibull distribution with use of a maximum likelihood technique. This approach suggests that the mean incubation period of paediatric TA‐AIDS cases is 2.4 years with a 90 per cent confidence interval ranging from 1.5 years to 7.2 years. The mode is 1.25 years. These estimates are much shorter than those recently published for adults, therefore the success of TA‐AIDS prevention strategies may be seen sooner in the paediatric age group.