Prediction of kidney cancer incidence in urban China by age-period-cohort Bayesian model

Objective To predict the disease burden of kidney cancer and to provide basic information for etiology and control planning.Methods We retrieved incidence data of kidney cancer from 18 urban cancer registries from National Central Cancer Registry during ten years period from 1998 to 2007.Ageperiod-cohort Bayesian model was applied for modeling to predict kidney cancer incidence in urban China in 2008-2015.Results Between 1998 and 2007,the incidence of kidney cancer in urban registration areas kept increasing dramatically.Incidence for male raised from 3.12/100 000 in 1998 to 5.36/100 000 in 2007 and from 1.66/100 000 to 2.67/100 000 for female.Different models showed that the increase was mainly caused by a cohort effect (P ＜ 0.001).The predicted incidence rate of kidney cancer for the year 2015 is 9.93 per 100 000 in male and 4.54 per 100 000 in female.The number of new cases will rise to 52 259 in 2015,including 36 616 men and 15 643 women.Conclusions The burden of kidney cancer in urban areas would increase due to the effect of age and cohort.Kidney cancer will become one of the main cancers threatening people's health in urban areas in China.Etiology research and planning of prevention and control for kidney cancer should be enhanced. Key words: Kidney neoplasms; Urban population; Bayesian analysis; Prediction