Objective  To analyze the incidence, clinical symptoms and pathogen detection results of infectious diarrhea in children under 5 years pld in Gansu province during 2009−2020, and predict the incidence of infectious diarrhea in the following 3 years by constructing autoregressive integrated moving average (ARIMA) model for the prevention and control of infectious diarrhea.

  Methods  The incidence data of infectious diarrhea in children under 5 years old in Gansu from 2009 to 2020 were collected from local sentinel surveillance hospitals for a descriptive epidemiological analysis. Software R4 1.2 was used to establish a seasonal ARIMA model for the prediction of the incidence of infectious diarrhea in following 3 years.

  Results  A total of 5 105 cases of infectious diarrhea were reported in children under 5 years old in Gansu from 2009 to 2020, there was no significant difference in incidence rate between boys and girls (P>0.05); The cases were mainly aged ≤2 years and the highest incidence rate was in age group 7−12 months. The annual cases showed a periodic distribution, with the peak in 2019. The cases were reported all the year round and the incidence peak was during June-November. Watery stool was one of the main clinical symptoms. The vomiting cases accounted for 42.38%, and the main character of vomiting was gastric content. About 1/3 of the sick children cases had fever. The positive rates of viral pathogen and bacterial pathogen were 50.16% (2012/4011) and 7.25% (271/3739) respectively. The main viral pathogens were rotavirus (33.15%), norovirus (14.29%) and adenovirus (8.32%) respectively, among which the detection rate of rotavirus A was highest. The main bacterial pathogens were diarrheagenic Escherichia coli (3.92%), Shigella (2.73%) and non-typhoid Salmonella (1.38%) respectively. The results of the seasonal ARIMA model forecast shows that the incidence in 2021-2023 is significantly lower than before, and then slowly rising after reaching the lowest point.

  Conclusion  The children under 2 years old were the population at high risk for infectious diarrhea. The infectious diarrhea was mainly caused by viral pathogens, which were predominated by rotavirus, and autumn and winter were the seasons with high incidence. The data predicted by the ARIMA (1,0,0) (1,0,0)₁₂ model can fit well with the incidence trend of infectious diarrhea in Gansu, which can be used for short-term prediction.