Runge-Ranzinger Silvia, Horstick Olaf, Marx Michael, Kroeger Axel
Abteilung Tropenhygiene & Offentliches Gesundheitswesen, Universitätsklinikum, Heidelberg, Germany.
Trop Med Int Health. 2008 Aug;13(8):1022-41. doi: 10.1111/j.1365-3156.2008.02112.x.
To review the evidence on the application of tools for dengue outbreak prediction/detection and trend monitoring in passive and active disease surveillance systems in order to develop recommendations for endemic countries and identify research needs.
Systematic review of literature in the Cochrane Database of Systematic Reviews, PubMed, EMBASE, Lilacs, WHO library database, manual reference search and grey literature. Two reviewers independently applied pre-defined inclusion and exclusion criteria and assessed the level of evidence. Studies describing the outcome of dengue disease surveillance with respect to trend monitoring and outbreak prediction/detection based on empirical data were included.
Twenty-four studies (of 1804 references) met the eligibility criteria. Different indicators and their respective threshold values were identified as potential triggers for outbreak alerts through retrospective analysis of data from passive and/or active surveillance systems. Some indicators are potentially useful for predicting imminent outbreaks in the low transmission season and others for detecting outbreaks at an early stage. However, the information collected is mainly retrospective and often site-specific and appropriate levels of sensitivity and specificity of the outbreak indicators/triggers could not be determined. Retrospective and prospective virus surveillance studies were not conclusive regarding the question of whether a newly introduced serotype is an outbreak predictor, but contributed additional indicators for outbreak prediction/detection. Under-reporting was a major concern. Taking cost and feasibility issues into account, it remains an open question whether dengue surveillance should be passive (based on routine reporting) or active (based on more costly sentinel or other active population based surveillance systems). Adding active surveillance elements to a well-functioning passive surveillance system improves sensitivity; adding laboratory elements to the system improves specificity.
In view of the lack of evidence about the most feasible and sustainable surveillance system in a country context, countries could use a stepwise approach to locally adapt their passive routine surveillance system into an improved combined active/passive surveillance approach. Prospective studies are needed to better define the most appropriate dengue surveillance system and trigger for dengue emergency response.
回顾在被动和主动疾病监测系统中应用登革热疫情预测/检测及趋势监测工具的证据，以便为流行国家制定建议并确定研究需求。
对Cochrane系统评价数据库、PubMed、EMBASE、Lilacs、世卫组织图书馆数据库中的文献进行系统评价，进行手动参考文献检索并查阅灰色文献。两名评审员独立应用预先定义的纳入和排除标准并评估证据水平。纳入基于经验数据描述登革热疾病监测中趋势监测及疫情预测/检测结果的研究。
1804篇参考文献中的24项研究符合纳入标准。通过对被动和/或主动监测系统的数据进行回顾性分析，确定了不同指标及其各自的阈值作为疫情警报的潜在触发因素。一些指标可能有助于预测低传播季节即将发生的疫情，另一些指标则有助于早期检测疫情。然而，所收集的信息主要是回顾性的，且往往针对特定地点，无法确定疫情指标/触发因素的适当敏感性和特异性水平。回顾性和前瞻性病毒监测研究对于新引入的血清型是否为疫情预测指标尚无定论，但为疫情预测/检测提供了额外指标。漏报是一个主要问题。考虑到成本和可行性问题，登革热监测应采用被动（基于常规报告）还是主动（基于成本更高的哨点或其他基于人群的主动监测系统）仍未确定。在运行良好的被动监测系统中增加主动监测要素可提高敏感性；在系统中增加实验室要素可提高特异性。
鉴于缺乏关于一个国家最可行和可持续监测系统的证据，各国可采用逐步方法，在当地将其被动常规监测系统调整为改进的主动/被动联合监测方法。需要进行前瞻性研究，以更好地确定最合适的登革热监测系统以及登革热应急反应的触发因素。
