目的应用乘积季节自回归移动平均(seasonal autoregressive integrated moving average,SARIMA)模型对肺结核发病率进行预测研究,探讨其可行性并为肺结核病的防治工作提供科学依据。方法应用EViews 7.0.0.1软件对我国2004年1月至2012年12月的肺结核逐月发病率建立乘积SARIMA模型并进行拟合,选取2013年1月至12月肺结核发病率数据评价模型的预测性能。结果建立的SARIMA(2,0,2)×(0,1,1)12模型能较好地拟合既往时间段内肺结核的发病率,对2013年1月至12月肺结核发病率的预测与实际发病率趋势基本吻合,平均误差绝对值为0.416 992,平均误差绝对率为5.350 8%。结论乘积SARIMA模型能较好地模拟和预测肺结核发病率在时间序列上的变动趋势,将其应用于肺结核发病预测是可行的,具有推广应用前景。 Objective To predict the incidence of pulmonary tuberculosis using seasonal autoregressive integrated moving average (SARIMA) model and explore its feasibility and provide a scientific basis for the prevention and treatment of tuberculosis. Methods EViews 7.0.0.1 software was used to establish the product SARIMA model of monthly incidence of pulmonary tuberculosis from January 2004 to December 2012 in our country. The predictive performance of the pulmonary TB incidence data evaluation model from January to December 2013 was evaluated. . Results The SARIMA (2, 0, 2) × (0, 1, 1) 12 model, which was established by the results, could well fit the incidence of tuberculosis in the past time period. The prediction of the incidence of tuberculosis in January-December 2013 and The trend of the actual incidence of the basic agreement, the average absolute value of the error was 0.416992, the average absolute error rate of 5.3508%. Conclusion The product SARIMA model can better simulate and predict the trend of time-series changes in the incidence of pulmonary tuberculosis. It is feasible to apply it to predict the incidence of pulmonary tuberculosis and has the prospect of popularization and application.