Time series analysis of mortality rate in Chongqing urban area

  Objective   To develop the health policy scientifically, the time series model is used to analyze and forecast the dynamic trend of the mortality rate in Chongqing urban area and explore the application of time series model in the medical field.

  Methods   Samples of population and death are taken out from the three communities in Chongqing urban area from 1994 to 2000 by cluster sampling.Combined with certain type time series recount multiplicity model and random type ARMA model, the time series model of the death rate is established in Chongqing urban area.

  Results   Determinant coefficient of the model(R²)is 0.743 5, Mean Absolute Percent Error(MAPE)is 10.59%.There is a rise in the mortality rate in Chongqing urban area following the decline in 1994, which falls gradually after 1998.The forecast shows that the mortality rate would decrease continually in 2001.

  Conclusion   Time series model can reveal the regularity of the mortality rate development objectively.The goodness of time series analysis and forecasting model is good, so it can be used to analyzed the regularity of disease or death variation and fore cast its future development trend.The effect of health care and health policy in recent years in Chongqing urban area is relatively good.