关键词: 肺结核, 自回归移动平均模型, 疫情, 预测

Abstract: Objective To establish an autoregressive integrated moving average (ARIMA) model to predict the monthly incidence of tuberculosis in Bao'an District, Shenzhen, and to provide a scientific reference for the formulation of tuberculosis prevention and control measures. Method The number of patients reported monthly in Bao'an District, Shenzhen from 2006 to 2020 was derived from the tuberculosis information management system of the China Disease Control and Prevention System. Use IBM SPSS 20.0 statistical software to establish a time series based on the number of patients reported monthly from January 2006 to December 2019, construct an ARIMA model, and screen out the optimal model through identification, ranking, and diagnosis of the model. The model is used to predict the monthly incidence of tuberculosis from 1 to 12 in 2020, and the prediction effect of the fitted model is evaluated by comparing the predicted value with the actual value. Result The parameters of the ARIMA(1,0,0)(0,1,1)₁₂ model passed the statistical test (P<0.05), and the residual sequence was a white noise sequence (P>0.05, R²=0.561, RMSE=22.632, NBIC=6.336), the goodness of fit is relatively good. The predicted value from January to December 2020 is basically consistent with the actual value, and the actual value falls within the 95% CI, and the prediction effect is good. Conclusion The ARIMA(1,0,0)(0,1,1)₁₂ model can be used for short-term prediction of tuberculosis epidemic in Bao'an District, Shenzhen, and the prediction effect is good.

Key words: Tuberculosis, Autoregressive integrated moving average model, Epidemic, Forecast