| Objective:To analyze the trend of incidence and mortality of stomach cancer in China during 2000-2014 using data from 22 cancer registries which had continuous surveillance data in the study period.To predict the nationwide incidence and mortality of stomach cancer in 2015,2020 and 2025 based on different models.Methods:1.Log-linear models were fitted for incidence and mortality rates,age-standardized incidence rates(ASIRs)and age-standardized mortality rates(ASMRs),and numbers of new cases and deaths in the 22 registries in 2000-2014 for total,male and female population.ASIR and ASMR were calculated based on Segi’s population.Trends of the indices were described by annual percent change(APC),average annual percent change(AAPC)and the 95%confidential intervals.2.Data were categorized into 18 age groups(0-4 years old,5-9 years old and so on to 80-84 years old,and 85 years old and above)for each gender.Incidence and mortality rates for each age group were calculated and fitted with Joinpoint regression model.To those age groups with low rates and statistically insignificant log-linear slopes,average rates were calculated as the predicted value of incidence and mortality rates in the future.To other age groups,the log-linear models of the latest period were used as the predict model to estimate the incidence and mortality rates in each age group in 2015,2020 and 20253.The male and female incidence and mortality rates in all age groups during 2000-2014 were treated as time series.Number of differencing to make the series stationary,and the orders of autoregressive and moving average for each time series were calculated to fit autoregressive integrated moving average(ARIMA)model Incidence and mortality rates in 2015,2020 and 2025 in each age group were estimated based on ARIMA models.4.The trends of male and female incidence and mortality rates in all birth cohorts and periods were analyzed.Using Bayesian Age-period-cohort Modeling and Prediction(BAMP),iterations were done for incidence and mortality rates by Markov Chain Monte Carlo method to estimate the incidence and mortality rates in 2015,2020 and 2025 in each age group.5.After comparing the estimated value for nationwide incidence and mortality rates of stomach cancer in each age group in 2013 and 2014 with the observed rates in the 22 cancer registries,the average ratio of these two values in the two years was calculated as correction factors to correct the difference between the nationwide rates and observed rates in each age group.Therefore,the nationwide incidence and mortality rates in each age group in 2015,2020 and 2025 could be estimated based on the results of the three models.Together with the population projection by United Nations,nationwide male and female incidence and mortality in the 3 years could be estimated,so that the burden of stomach cancer in the total,male and female population can be calculated.Results:1.Incidence rate of stomach cancer in the 22 cancer registries remained stable in 2000-2014(AAPC=0.2%),while ASIR decreased significantly(AAPC=-2.9%)and number of new cases increased significantly(AAPC=2.8%).In the same period,mortality rate remained stable(AAPC=-0.5%),while ASMR decreased significantly(AAPC=-3.8%)and number of deaths increased significantly(AAPC=2.3%).2.According to Joinpoint regression model,incidence rate of stomach cancer will rise from 28.00 per 100,000 in 2015 to 29.31 per 100,000 in 2025.ASIR will drop from 19.91 to 15.88 per 100,000,number of new cases with rise from 391,200 to 421,700,and the proportion of elderly patients(60 years old and above)will rise from 71.0%to 77.7%in 2015-2025.Mortality rate of stomach cancer will drop from 19.20 per 100,000 in 2015 to 18.78 per 100,000 in 2025.ASMR will drop from 13.50 to 9.88 per 100,000,number of deaths will rise from 268,200 to 270,300,and the proportion of elderly patients will rise from 78.3%to 83.2%in 2015-2025.3.According to ARIMA model,incidence rate of stomach cancer will drop from 27.93 per 100,000 in 2015 to 27.66 per 100,000 in 2025.ASIR will drop from 19.79 to 15.05 per 100,000,number of new cases with rise from 390,200 to 397,900,and the proportion of elderly patients will rise from 70.8%to 72.0%in 2015-2025.Mortality rate of stomach cancer will drop from 18.74 per 100,000 in 2015 to 13.39 per 100,000 in 2025.ASMR will drop from 13.19 to 6.99 per 100,000,number of deaths will drop from 262,000 to 193,000,and the proportion of elderly patients will rise from 78.5%to 81.0%in 2015-2025.4.According to BAMP,incidence rate of stomach cancer will drop from 27.18 per 100,000 in 2015 to 22.70 per 100,000 in 2025.ASIR will drop from 19.30 to 12.23 per 100,000,number of new cases will drop from 379,700 to 326,600,and the proportion of elderly patients will rise from 70.4%to 76.5%in 2015-2025.Mortality rate of stomach cancer will drop from 17.75 per 100,000 in 2015 to 8.81 per 100,000 in 2025.ASMR will drop from 12.50 to 4.59 per 100,000,number of deaths will drop from 248,000 to 126,800,and the proportion of elderly patients will rise from 78.2%to 82.6%in 2015-2025.Conclusion:The ageing and increasing of population may be the main reason for the increased burden of stomach cancer in China during 2000-2014.Despite of certain diversities in the results of the three models,they all showed that incidence and mortality rate of stomach cancer in China will not increase dramatically,ASIR and ASMR will decrease and the proportion of elderly patients will increase in China in 2015-2025.In the three models,Joinpoint regression model is the most conservative and suitable for short and mid-term prediction.The accuracy of ARIMA decreased as the forecast prolonged,making ARIMA suitable for only short-term prediction.BAMP is the most radical in the declining trend of stomach cancer burden.The results of BAMP remained accurate as the forecast prolonged,making BAMP a better choice for mid and long-term prediction. |
PDF Full Text Request