Background: Hemorrhagic fever with renal syndrome (HFRS), a rodent-borne disease caused by different species of hantaviruses, is widely endemic in China. Shandong Province is one of the most affected areas. This study aims to analyze the epidemiological characteristics of HFRS, and to predict the regional risk in Shandong Province.
Methods: Descriptive statistics were used to elucidate the epidemiological characteristics of HFRS cases in Shandong Province from 2010 to 2018. Based on environmental and socioeconomic data, the boosted regression tree (BRT) model was applied to identify important influencing factors, as well as predict the infection risk zones of HFRS.
Results: A total of 11,432 HFRS cases were reported from 2010 to 2018 in Shandong, with groups aged 31-70 years (81.04%), and farmers (84.44%) being the majority. Most cases were from central and southeast Shandong. There were two incidence peak periods in April to June and October to December, respectively. According to the BRT model, we found that population density (a relative contribution of 15.90%), elevation (12.02%), grassland (11.06%), cultivated land (9.98%), rural settlement (9.25%), woodland (8.71%), and water body (8.63%) were relatively important influencing factors for HFRS epidemics, and the predicted high infection risk areas were concentrated in central and eastern areas of Shandong Province. The BRT model provided an overall prediction accuracy, with an area under the receiver operating characteristic curve of 0.91 (range: 0.83-0.95).
Conclusions: HFRS in Shandong Province has shown seasonal and spatial clustering characteristics. Middle-aged and elderly farmers are a high-risk population. The BRT model has satisfactory predictive capability in stratifying the regional risk of HFRS at a county level in Shandong Province, which could serve as an important tool for risk assessment of HFRS to deploy prevention and control measures.
Keywords: boosted regression tree model; epidemiological characteristics; hemorrhagic fever with renal syndrome; influencing factors; regional risk prediction.