Abstract

Objectives: Based on observed numbers of incident mesotheliomas since 1972, to predict future numbers in men in New South Wales.

Methods: The incidence of mesothelioma was modelled in two ways. First by using an age/birth cohort model, and second by using a model based on potential exposure to asbestos in terms of age and calendar year. The latter model included a term for clearance of asbestos fibres from the lungs, and a term for diagnostic fraction. The age and calendar year model was based on the model introduced by Hodgson and colleagues but replaced piecewise effects by smooth functions represented by cubic splines.

Results: The number of mesotheliomas between 2004 and 2060 was predicted as 6690 with the age-cohort model and as 6779 by the age and calendar year model, with peak annual numbers of 187 in the year 2021 and 196 in the year 2014 with the two models respectively.

Conclusions: The pattern of parameter estimates in the two models was in accord with the known use of amphibole asbestos in Australia. The predicted peak year of 201421 is 3035 years after the phasing out of amphibole use, and this period is in accord with predictions for the UK and the US; in the latter country the peak was 1015 years earlier corresponding to a marked decline of amphibole use in and following the 1960s.