重庆市万州区结直肠癌死亡趋势分析

Analysis on colorectal cancer mortality trends based on GM(1,1)model in Wanzhou district of Chongqing city

目的 GM(1,1)模型是建立在灰色系统理论之上,以时间序列为研究对象,去发现事物发展变化的、连续的和离散的未来时间序列,从而找出事物发展中蕴含的规律。本研究采用Matlab软件建立GM(1,1)数学模型对2005-2014年重庆市万州区结直肠癌死亡情况进行描述,并预测未来3年结直肠癌死亡率,为疾病防治提供参考依据。方法收集万州区全部52个乡、镇和街道2005-2014年居民恶性肿瘤死亡登记资料,根据结直肠癌死亡数据建立结直肠癌死亡率GM(1,1)灰色模型,并预测2015-2017年本地区结直肠癌的死亡率。结果计算出本区结直肠癌死亡率GM(1,1)模型为X(t)=exp[119.902e^(0.016(t-1))-117.958],整体拟合平均误差为5.17%,后验差比值C=0.34,小误差概率P=1,模型精度为Ⅰ级(优)。预测结直肠癌死亡率2015年为10.44/10万,2016年为10.86/10万,2017年为11.30/10万。结论预测模型达到"优"等级,预测未来3年的结直肠癌死亡率呈上升趋势。

OBJECTIVE A mathematical model by Matlab of colorectal cancer death Wanzhou District of Chongqing was described from 2005 to 2014,and predicted colorectal cancer deaths over the next three years,to provide reference for disease control and prevention.METHODS The registered data of cancer deaths were collected in all 52 townships,towns and streets in Wanzhou district from 2005 to 2014,establishing colorectal cancer mortality GM（1,1）gray model based on colorectal cancer mortality data,the accuracy of the model was tested,and the mortality rate of colorectal cancer from 2015 to 2017in the region was predicted.RESULTS The model of colorectal cancer mortality GM（1,1）in the area was X（t）=exp[119.902e^（0.016（t-1））-117.958],the average relative error was 5.17%,posterior variance ratio C was 0.34,small error probability P was 1,and model accuracy was grade I（excellent）.Predicting colorectal cancer mortality rate was 10.44/100 000 in 2015,10.86/100 000 in 2016,and 11.30/100 000 in 2017.CONCLUSION The predictive model reached the excellent level,and colorectal cancer mortality in the next three years is on the rise.