[Progress in research of rash and fever syndrome surveillance and early warning].

Abstract

Objective: To introduce the progress in research of rash and fever syndrome (RFS) surveillance and early warning both at home and abroad, and provide reference for surveillance and prevention of RFS in China. Methods: The keywords "fever" "rash" and "surveillance" and others were used for a literature retrieval by using China National Knowledge Infrastructure, Wanfang Data Knowledge Service Platform, PubMed and Web of Science. The languages of literatures were limited in Chinese and English. The key information of the literatures were collected and analyzed with Excel. Results: A total of 36 study papers (21 in Chinese and 15 in English) were included. The studies mainly focused on the pathogen surveillance of RFS (n=19). The pathogens included measles virus, varicella-zoster virus, rubella virus, enterovirus, human B19 virus, dengue virus, streptococcus group A, Salmonella typhi and Salmonella paratyphoid,human herpesvirus, mumps virus and adenovirus. Eight studies were about the surveillance in major events, such as sport game, World Expo and religious gathering, or sudden natural disasters, such as earthquake and tropical storm, during 2010-2015. Eight studies focused on case or epidemic surveillance, most of which were studies from other counties. The surveillance sites were medical institutions. RFS was diagnosed according to the International Classification of Diseases, 9th (ICD-9) and symptoms descripted in chief-complaint. Only one study in Mongolia conducted RFS epidemic prediction. The analysis methods of 36 papers included simple descriptive analysis, time-based early warning models (such as regression analysis, fixed threshold method, Hugh Hart control chart method and cumulative sum control chart method) and time series analysis method. Conclusions: In the future, RFS surveillance system should cover both known pathogens and emerging pathogens. Automatic surveillance using information capture and intelligent modelling can be applied to improve the sensitivity and specificity of RFS surveillance and early warning.