Pivette Mathilde, Mueller Judith E, Crépey Pascal, Bar-Hen Avner
EHESP French School of Public Health, Sorbonne Paris Cité, Rennes, France.
Université Paris Descartes, MAP5, Paris, France.
BMC Infect Dis. 2014 Nov 18;14:604. doi: 10.1186/s12879-014-0604-2.
This systematic literature review aimed to summarize evidence for the added value of drug sales data analysis for the surveillance of infectious diseases.
A search for relevant publications was conducted in Pubmed, Embase, Scopus, Cochrane Library, African Index Medicus and Lilacs databases. Retrieved studies were evaluated in terms of objectives, diseases studied, data sources, methodologies and performance for real-time surveillance. Most studies compared drug sales data to reference surveillance data using correlation measurements or indicators of outbreak detection performance (sensitivity, specificity, timeliness of the detection).
We screened 3266 articles and included 27 in the review. Most studies focused on acute respiratory and gastroenteritis infections. Nineteen studies retrospectively compared drug sales data to reference clinical data, and significant correlations were observed in 17 of them. Four studies found that over-the-counter drug sales preceded clinical data in terms of incidence increase. Five studies developed and evaluated statistical algorithms for selecting drug groups to monitor specific diseases. Another three studies developed models to predict incidence increase from drug sales.
Drug sales data analyses appear to be a useful tool for surveillance of gastrointestinal and respiratory disease, and OTC drugs have the potential for early outbreak detection. Their utility remains to be investigated for other diseases, in particular those poorly surveyed.
本系统文献综述旨在总结药物销售数据分析在传染病监测方面附加价值的证据。
在PubMed、Embase、Scopus、Cochrane图书馆、非洲医学索引和Lilacs数据库中检索相关出版物。根据研究目的、所研究疾病、数据来源、方法以及实时监测性能对检索到的研究进行评估。大多数研究使用相关性测量或疫情检测性能指标（敏感性、特异性、检测及时性）将药物销售数据与参考监测数据进行比较。
我们筛选了3266篇文章，纳入本综述的有27篇。大多数研究聚焦于急性呼吸道和胃肠道感染。19项研究回顾性地将药物销售数据与参考临床数据进行比较，其中17项观察到显著相关性。4项研究发现非处方药销售在发病率增加方面先于临床数据。5项研究开发并评估了用于选择药物组以监测特定疾病的统计算法。另外3项研究建立了从药物销售预测发病率增加的模型。
药物销售数据分析似乎是胃肠道和呼吸道疾病监测的有用工具，非处方药有早期疫情检测的潜力。其在其他疾病，尤其是监测不足的疾病中的效用仍有待研究。
