Epidemiological and time series analysis of haemorrhagic fever with renal syndrome from 2004 to 2017 in Shandong Province, China

Abstract

Shandong Province is an area of China with a high incidence of haemorrhagic fever with renal syndrome (HFRS); however, the general epidemic trend of HFRS in Shandong remains unclear. Therefore, we established a mathematical model to predict the incidence trend of HFRS and used Joinpoint regression analysis, a generalised additive model (GAM), and other methods to evaluate the data. Incidence data from the first half of 2018 were included in a range predicted by a modified sum autoregressive integrated moving average-support vector machine (ARIMA-SVM) combination model. The highest incidence of HFRS occurred in October and November, and the annual mortality rate decreased by 7.3% (p < 0.05) from 2004 to 2017. In cold months, the incidence of HFRS increased by 4%, −1%, and 0.8% for every unit increase in temperature, relative humidity, and rainfall, respectively; in warm months, this incidence changed by 2%, −3%, and 0% respectively. Overall, HFRS incidence and mortality in Shandong showed a downward trend over the past 10 years. In both cold and warm months, the effects of temperature, relative humidity, and rainfall on HFRS incidence varied. A modified ARIMA-SVM combination model could effectively predict the occurrence of HFRS.

Effects and interaction of meteorological factors on hemorrhagic fever with renal syndrome incidence in Huludao City, northeastern China, 2007–2018

“…Consistent with previous studies, our study found that rising temperatures increased the risk of HFRS infection [13,34]; for example, Liu et al found that temperatures between 10-25 degrees Celsius was a favorable condition for HFRS transmission in Junan County in Shandong [14]. Xiang et al pooled the results of China's 19 cities and showed that a 1˚C increase in temperature resulted in a 1.6% (95% CI, 1.0%-2.2%) increase in HFRS [35].…”

“…Previous studies have researched the associations between climatic factors and HFRS epidemic risks [12][13][14]. Zhang et al found that monthly average air temperature was nonlinearly correlated with the monthly incidence of HFRS and reached the highest relative risk (RR) at approximately 23 degrees Celsius in Shandong Province [15]. A study carried out in Guangzhou indicated that lags in temperature from 1-3 months, rainfall of 2 months and relative humidity of 4 months all have significant associations with the incidence of HFRS.…”

Misdiagnosis of scrub typhus as hemorrhagic fever with renal syndrome and potential co-infection of both diseases in patients in Shandong Province, China, 2013–2014

“…This led to physicians commonly diagnosing febrile patients with atypical symptoms as HFRS. Because all three aforementioned diseases are prevalent in Shandong Province [15][16][17], the possibility of misdiagnosis and co-infection cannot be ignored, and further epidemiological investigation is needed. Except for confirmed scrub typhus, SFTS and HFRS cases, a considerable number (n = 45, 35%) of patients were still unconfirmed PLOS NEGLECTED TROPICAL DISEASES although they were clinically diagnosed as HFRS.…”

Time Series Analysis of Hemorrhagic Fever with Renal Syndrome in Mainland China by using XGBoost Forecasting Model

“…After combining the defined loss function and complexity of the tree, the objective function can be expressed by formula (13). by the frequency functions used to split the feature.…”

“…At present, some models had been used in predicting HFRS, including neural networks [12], and general addictive model (GAM), etc. [13]. Most of these methods are based on one-step forecasting.…”