| Objective:Comprehensively,to grasp the injury death level,life lost and its changed trend of residents in the XINJIAN district from Nanchang city from 2011 to2015,in order to predict the trend of injury death in the next three years,we plan to establish the ARIMA model.So as to provide scientific basis for health management decision-making,put forward countermeasures and suggestions to prevent and control death by injury,ultimately,improving the people’s life quality and health level.Methods:All the death cases of residents from the National Death Report Information System by Census method,also,we combined with the results of investigation of missing reports;population data was provided by the Nanchang statistics.The classification of injuries were based on the International Classification of Diseases(ICD-10).Descriptive analysis was used to analyze the mortality,death composition,and death cause sequence;simple life and death life table analysis based on SPSS 20.0 software to establish programming language,potential years of lost life and life lost rate calculation by Class Midpoint method;ARIMA model was established by Eviews 8.0 software,in order to predict the number of injury deaths per month for residents in the next three years,the right model was obtained by the sequence smoothing,model identification,model parameter estimation and adaptability test.Results:From 2011 to 2015,the deaths of XINJIAN district were 18693,the crude death rate was 538.38/10 million,the standardized mortality rate was 563.00/10 million,the cause of death in the top five were follows:circulatory system diseases,cancer,respiratory system diseases,injury,endocrine,nutrition and metabolic diseases;the life expectancy of residents was 78.38 years old,respectively,74.71 years old for men and 83.04 years old for women;the potential years of life lost of the whole-population was 46191.5 years,the potential years of life lost rate was14.26‰,average years of life lost was 33.14 years,these were ranked first;standardized potential years of life lost was 44774.34 years,the standardizedpotential years of life lost rate was 12.90‰,these were ranked second,different sex residents injury life lost index the same with whole-population;the number of injury death were 1639,including 1106 males and 533 females;the mortality rate was38.36~53.72/10 million from 2011 to 2015,and degree of declining was13.88%,EAPC was-2.05%,in the previous three years,injuries were ranked third,were ranked fourth from 2014 to 2015;for the composition ratio,the injury ratio was increased from 8.62% in 2011 to 9.95% in 2013,and was declined from 8.44% in2014 to 7.45% in 2015;the mortality rate of males was higher than females(χ2 =25.747,P<0.05),sex ratio range from 1.81 to 2.34;except in 2012,in 2013,injury mortality were statistically difference between males and females;0~ age group residents mortality were higher,reach to 20~ age group were 24.66/10 million decline to a minimum;with the increase of age,the injury mortality rate was generally increased,the male lowest peak was appeared in the 25~ age group,while the women was appeared in the 10~ age group;the composition ratio in three categories of injury death as follows:accidental injury accounted for 89.51%,intentional injury accounted for 4.15%,intentions of unknown damage accounted for 4.53%,crude mortality rates were 42.25/10 million,3.89/10 million,1.07/10 million;annually,the proportion of accidental damage was highest(86.65~92.97%),the lowest in 2014 was 86.65%,and the highest in 2012 was 92.97%;among the top five injuries were accidental,accidental drowning,suicide,accidental fall,accidental poisoning,accounting for49.48%,17.69%,7.38%,6.35% and 3.17% respectively,accounting for 84.08% in total,the first two causes of injury death were same in each year;the potential life loss to injury were 79.90 years old,the males were 76.44 years old,females were 84.26 years old,the whole-population were biggest lost 1.74 years due to injury in 2013;for potential life loss in three categories of injury death,accidental injury were highest,respectively,the males were 76.31 years old,females were 84.12 years old;for potential life loss in fifth causes of injury death,the traffic accident were 79.09 years old,the accidental drowning death were 78.75 years old,the suicide were 78.47 years old,the accidental fall were 78.46 years old,the accidental poisoning were 78.42 years old,males and females were slightly different,also,the top two were traffic accidents and accidental drowning;from January 2011 to December 2015,the monthly data ofinjury in Nanchang XINJIAN district showed a long-term downward trend,a peak in August each year,and didn,t pass ADF test,suggested that the original sequence was non-stationary sequence;the first order difference and the first order seasonal difference were applied to the original sequence,the results showed that the ADF statistic was-11.68,which had good stability and conformed to the modeling conditions;after the model recognition,the auto-correlation coefficient and the partial correlation coefficient all showed the first order truncation,we considered to establish model of multiple seasonal ARIMA(p,1,q)×(P,1,Q)12;according to the least squares principle,we compared R2,AIC,SC values from statistically significant model in parameters,finally,we got a best model of multiple seasonal ARIMA(0,1,1)×(0,1,1)12,that moment R2=0.64,AIC=6.82,SC=6.89;the residual sequence was verified by white noise test,model of multiple seasonal ARIMA(0,1,1)×(0,1,1)12 was used to predict the number of monthly deaths from January 2016 to December 2018 in XINJIAN district from Nanchang city,the average absolute error of the model was5.49%,the hill unequal coefficient was 0.12,the deviation rate was 0.07,and the variance rate was 0.19.Conclusion:From2011 to 2015,the injury was at the fourth place,had always been an important reason for the residents in XINJIAN district.Although the injury rate showed a downward trend,the life loss was ranked first by injury death among all deaths;three categories of injury death,due to accidental damage lead to the highest in composition ratio,mortality and potential life loss,so,we should focus on accidental injury;for type of injury,health managers should focus on raffic accidents,accidental drowning,suicide,accidental fall,accidental poisoning;ARIMA(0,1,1)(0,1,1)12 model had good prediction accuracy and can be used to predict the short-term trend of the monthly death toll,ultimately,it can provide an important basis for the disease prevention and control departments by grasping the rules of injury,and the evaluating the effects after the implementation of preventive measures. |
PDF Full Text Request