Using Lee-Carter Model to Fit and Forecast Age-Specific Mortality Rate for Colon Cancer

Abstract

This study was mainly focused on modeling and forecasting cancer mortality rates using standard Lee Carter Model, This model has become reference and leading statistical model for forecasting mortality. The model is applied to Colon cancer mortality data for Egyptian for both sexes ((male-female) aged from 5 to 74 years for the time period 2001-2014 and forecast mortality index by ARIMA Random Walk with Drift (0,1,0) from 2015 to 2020. The data obtained from World Health Organization (WHO).The model's parameters estimated by Singular Value Decomposition (SVD) and Maximum Likelihood Estimation (MLE), the comparison of the two methods (SVD, MLE) based on the mean error (ME) and mean square error (MSE). The results obtained by using different statistics packages like R and Iterative Lee Carter (ilc). The results showed MLE is better to estimate the parameters (ME=0.00506, MSE=0.11065) for male, while for female the SVD is better to estimate the parameters (ME=0.00835, MSE=0.07022). The error of forecasting age-specific mortality rate was (MPE=0.30682) for male and the error of forecast age-specific cancer mortality rate (MPE=0.04833) for female. The results showed the younger ages from 5 to 13, and from 30 to 39 had lower cancer mortality rate for male, and higher age-specific cancer mortality rate was in age-group (70-74) and it was 27.91 in year 2020 (per 100,000) for male .While for female had lower age-specific cancer mortality rate in younger ages from 5 to 24 and higher age-specific cancer mortality rate in old ages especially (60-64) and it was 21.55 in year 2020 (per 100,000). The study came out with a number of recommendation, the most important were to use Lee-Carter method to modeling and forecasting age-specific mortality rate , and to have care and accuracy when registering data of died people.