Objective To establish an infectious disease index forecast model in Zhuhai city to provide ideas for the forecast of infectious disease risk. Methods The lower, midline and upper control limits of Statistical Process Control(SPC) were used to divide the city′s 2014-2017 influenza-like case ratio, hand-foot-mouth disease and other infectious diarrhea risk grades in the weekly measurement unit from 2014 to 2017 (Breteau index was judged by 5, 10, 20). Long and short-term memory neural network model (LSTM) and autoregressive integrated moving average model (ARIMA)were used to predict the data of 15-19 weeks in 2018. The infectious disease index was calculated and compared the predicted value with the actual value to evaluate the prediction consistency. Results In the LSTM model of hand, foot and mouth disease incidence in Zhuhai, the MSE of the test set was 9.0441, the RMSE was 3.0073, the MSE of the training set was 1.1812, and the RMSE was 1.0868. The remaining models performed well on the training set and test set, and there was no overfitting phenomenon. Comparing the risk index level prediction with the actual value, the prediction agreement rate was 96.0%. Conclusion It was feasible to use SPC to classify risk levels and to use LSTM to construct infectious disease index prediction models.