文献速递免疫规划传染病防控与疫情报告

基于树-时间扫描分析的水痘疫苗安全性监测

Safety surveillance of varicella vaccine using tree-temporal scan analysis.
PubMed 文献 · Vaccine · 2021-09-21

本研究利用台湾全民健康保险数据库和免疫信息系统2004-2014年数据,对1,194,189名12-35月龄接种首剂水痘疫苗的儿童进行自我对照树-时间扫描分析,识别接种后1-56天内不良事件聚集。结果发现发热、胃炎十二指肠炎、急性上呼吸道感染和水痘感染四个已知安全信号,未检测到非预期聚集,表明该方法用于幼儿疫苗安全性监测可行。

要点

  • 水痘疫苗首剂接种后未发现非预期不良事件聚集。
  • 识别出的四个信号均为已知安全特征,与先前监测一致。
  • 树-时间扫描统计方法适用于幼儿疫苗安全性监测。

核心结论

  • 大规模儿童水痘疫苗接种后未检测到非预期不良事件聚集,树-时间统计方法可行。

研究设计与样本

  • 全国性安全性监测,基于台湾全民健康保险数据库和免疫信息系统2004-2014年数据,纳入12-35月龄接种首剂水痘疫苗儿童1,194,189名,采用自我对照树-时间数据挖掘工具分析接种后1-56天内的ICD-9-CM诊断。

主要发现

  • 发热在接种后1-6天聚集,归因风险38.5/10万,p<0.001。
  • 胃炎和十二指肠炎在1-2天聚集,归因风险5.9/10万,p<0.001。
  • 急性上呼吸道感染在1-5天聚集,归因风险11.0/10万,p=0.006。
  • 水痘感染在1-9天聚集,归因风险2.7/10万,p<0.001。

临床怎么用

  • 接种水痘疫苗后1-9天需关注发热、上呼吸道感染、胃肠道症状及水痘感染,但均为已知且罕见信号。
  • 树-时间扫描可作为疫苗安全性主动监测的补充工具。

局限性

  • 被动监测系统可能存在不良事件漏报和接种人群信息缺乏。
  • 本研究仅纳入12-35月龄儿童首剂水痘疫苗,结论外推需谨慎。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

IMPORTANCE: Passive surveillance systems are susceptible to the under-reporting of adverse events (AE) and a lack of information pertaining to vaccinated populations. Conventional active surveillance focuses on predefined AEs. Advanced data mining tools could be used to identify unusual clusters of potential AEs after vaccination. OBJECTIVE: To assess the feasibility of a novel tree-based statistical approach to the identification of AE clustering following the implementation of a varicella vaccination program among one-year-olds. SETTING AND PARTICIPANTS: This nationwide safety surveillance was based on data from the Taiwan National Health Insurance database and National Immunization Information System for the period 2004 through 2014. The study population was children aged 12-35 months who received the varicella vaccine. EXPOSURE: First-dose varicella vaccine. OUTCOMES AND MEASURES: All incident ICD-9-CM diagnoses (emergency or inpatient departments) occurring 1-56 days after the varicella vaccination were classified within a hierarchical system of diagnosis categories using Multi-Level Clinical Classifications Software. A self-controlled tree-temporal data mining tool was then used to explore the incidence of AE clustering with a variety of potential risk intervals. The comparison interval consisted of days in the 56-day follow-up period that fell outside the risk interval. RESULTS: Among 1,194,189 varicella vaccinees with no other same-day vaccinations, nine diagnoses with clustering features were categorized into four safety signals: fever on days 1-6 (attributable risk [AR] 38.5 per 100,000, p < 0.001), gastritis and duodenitis on days 1-2 (AR 5.9 per 100,000, p < 0.001), acute upper respiratory infection on days 1-5 (AR 11.0 per 100,000, p = 0.006), and varicella infection on days 1-9 (AR 2.7 per 100,000, p < 0.001). These safety profiles and their corresponding risk intervals have been identified in previous safety surveillance studies. CONCLUSIONS: Unexpected clusters of AEs were not detected after the mass administration of childhood varicella vaccines in Taiwan. The tree-temporal statistical method is a feasible approach to the safety surveillance of vaccines in populations of young children.

来源:Vaccine,查看原文。版权归原作者所有。

引用

  1. Liu CH, Huang WT, Chie WC, et al. Safety surveillance of varicella vaccine using tree-temporal scan analysis. Vaccine. 2021. PMID: 34561139. doi:10.1016/j.vaccine.2021.09.035

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。