School of Public Health and Social Work, Institute of Health and Biomedical Innovation, Queensland University of Technology, Brisbane, Queensland, Australia.
Science and Engineering Faculty, Mathematical Sciences and Centre for Data Science, Queensland University of Technology, Brisbane, Queensland, Australia.
Int J Biometeorol. 2021 Dec;65(12):2203-2214. doi: 10.1007/s00484-021-02155-4. Epub 2021 Jun 1.
The use of internet-based query data offers a novel approach to improve disease surveillance and provides timely disease information. This paper systematically reviewed the literature on infectious disease predictions using internet-based query data and climate factors, discussed the current research progress and challenges, and provided some recommendations for future studies. We searched the relevant articles in the PubMed, Scopus, and Web of Science databases between January 2000 and December 2019. We initially included studies that used internet-based query data to predict infectious disease epidemics, then we further filtered and appraised the studies that used both internet-based query data and climate factors. In total, 129 relevant papers were included in the review. The results showed that most studies used a simple descriptive approach (n=80; 62%) to detect epidemics of influenza (including influenza-like illness (ILI)) (n=88; 68%) and dengue (n=9; 7%). Most studies (n=61; 47%) purely used internet search metrics to predict the epidemics of infectious diseases, while only 3 out of the 129 papers included both climate variables and internet-based query data. Our research shows that including internet-based query data and climate variables could better predict climate-sensitive infectious disease epidemics; however, this method has not been widely used to date. Moreover, previous studies did not sufficiently consider the spatiotemporal uncertainty of infectious diseases. Our review suggests that further research should use both internet-based query and climate data to develop predictive models for climate-sensitive infectious diseases based on spatiotemporal models.
利用基于互联网的查询数据为改善疾病监测和提供及时的疾病信息提供了一种新方法。本文系统地回顾了利用基于互联网的查询数据和气候因素预测传染病的文献，讨论了当前的研究进展和挑战，并为未来的研究提供了一些建议。我们在 PubMed、Scopus 和 Web of Science 数据库中检索了 2000 年 1 月至 2019 年 12 月之间的相关文章。我们最初纳入了使用基于互联网的查询数据预测传染病流行的研究，然后进一步筛选和评估了同时使用基于互联网的查询数据和气候因素的研究。共有 129 篇相关论文纳入综述。结果表明，大多数研究（n=80；62%）采用简单描述性方法（n=88；68%）来检测流感（包括流感样疾病（ILI））（n=88；68%）和登革热（n=9；7%）的流行。大多数研究（n=61；47%）纯粹使用互联网搜索指标来预测传染病的流行，而在 129 篇论文中只有 3 篇同时包含气候变量和基于互联网的查询数据。我们的研究表明，同时考虑基于互联网的查询数据和气候变量可以更好地预测对气候敏感的传染病流行，但到目前为止，这种方法尚未得到广泛应用。此外，以前的研究没有充分考虑传染病的时空不确定性。我们的综述表明，进一步的研究应该使用基于互联网的查询和气候数据，根据时空模型开发针对气候敏感传染病的预测模型。
