A SARIMA (1, 1, 2) (0, 1, 1)12 model accurately predicted the annual periodicity and seasonal variation of HFMD incidence in China, with a mean error rate of 16.86% and a determination coefficient of 94.27%.
Observational
The SARIMA (1, 1, 2) (0, 1, 1)12 model accurately captures the annual periodicity and seasonal variation of HFMD incidence in China, providing a useful tool for forecasting and public health planning.
Effect estimate: MER 16.86%, R2 94.27%
Seasonal autoregressive-integrated moving average (SARIMA) has been widely used to model and forecast incidence of infectious diseases in time-series analysis. This study aimed to model and forecast monthly cases of hand, foot and mouth disease (HFMD) in China. Monthly incidence HFMD cases in China from May 2008 to August 2018 were analysed with the SARIMA model. A seasonal variation of HFMD incidence was found from May 2008 to August 2018 in China, with a predominant peak from April to July and a trough from January to March. In addition, the annual peak occurred periodically with a large annual peak followed by a relatively small annual peak. A SARIMA model of SARIMA (1, 1, 2) (0, 1, 1)12 was identified, and the mean error rate and determination coefficient were 16.86% and 94.27%, respectively. There was an annual periodicity and seasonal variation of HFMD incidence in China, which could be predicted well by a SARIMA (1, 1, 2) (0, 1, 1)12 model.
Tian et al. (Tue,) conducted a observational in Hand, foot and mouth disease (HFMD). Seasonal autoregressive-integrated moving average (SARIMA) model was evaluated on Model accuracy (mean error rate and determination coefficient) (MER 16.86%, R2 94.27%). A SARIMA (1, 1, 2) (0, 1, 1)12 model accurately predicted the annual periodicity and seasonal variation of HFMD incidence in China, with a mean error rate of 16.86% and a determination coefficient of 94.27%.