Evaluation of hand-wash intervention on the incidence of hand, foot, and mouth disease in Guangzhou city using ARIMAX model

Abstract

Objective To evaluate the effect of hand-wash intervention on the incidence of hand, foot, and mouth disease (HFMD) in the children under 5 years old in Guangzhou by ARIMAX model. Methods The data of 447 976 HFMD cases and weekly meteorological data, including temperature, accumulated rainfall, relative humidity, and air pressure in Guangzhou city from 2008 to 2016 were collected. The ARIMAX model was constructed based on hand-wash intervention factor and meteorological data to evaluate the effect of hand-wash intervention on the incidence of HFMD. Results From 2008 to 2015, the median (quartiles) of HFMD cases per week was 559 (150~1 386). The weekly average air pressure was (1 009.10±6.57) hPa and weekly average value of relative humidity was (75.03±8.33) %. The ARIMAX model showed that incidence of HFMD one week afterward was statistically associated with air pressure (β=-2.44, t=-3.19, P=0.001), relative humidity (β=2.49, t=2.12, P=0.017) and hand-wash intervention (β=-180.18, t=-2.87, P=0.002) while AIC and RMSE were 4 102.34 and 198.65, respectively. In general, the predictive values by the model matched the actual values in 2016 with RMSE of 549.80. Conclusions The ARIMAX model predicted effectively the incidence of HFMD with controlled air pressure and relative humidity. Hand-wash intervention can reduce the case numbers of HFMD in one week later. Campaign of hand-wash intervention in May is recommended. Key words: Hand, foot, and mouth disease; Hand-wash; Linear models; Regression analysis; Health promotion