METHODS: We used self-controlled case series (SCCS) design to analyze data from US Veterans ≥18 years with coronary/cerebrovascular or exacerbation event +/-1 year of lab-confirmed influenza (LCI) during 2010-2018. We estimated the incidence ratio (IR) (95% CI) of the event for risk interval (Days 1-7 post-LCI) versus control interval (all other times +/-1 year of LCI) with fixed-effects conditional Poisson regression. We included biomarker data for mediation analysis.
RESULTS: We identified 3439 episodes with coronary/cerebrovascular-related hospitalizations. IRs (95% CI) for LCI risk versus control interval were STEMI 0.6 (0.1, 4.4), NSTEMI 7.3 (5.8, 9.2), ischemic stroke 4.0 (3.0, 5.4), hemorrhagic stroke 6.2 (3.4, 11.5), and coronary spasm 1.3 (0.5, 3.0). IR significantly increased for NSTEMI and ischemic stroke among those ≥ 65 years. IR for NSTEMI and ischemic stroke dropped 26% and 10%, respectively, when white blood cell (WBC) and platelet count were considered. LCI was significantly associated with exacerbation of preexisting asthma, chronic obstructive pulmonary disease, and congestive heart failure.
CONCLUSIONS: We found significant association between LCI and hospitalization for NSTEMI, ischemic stroke, and hemorrhagic stroke, the latter possibly due to unaccounted time-varying confounding in SCCS design.
方法:我们使用自我对照病例系列(SCCS)设计来分析来自美国退伍军人≥18岁的冠状动脉/脑血管或恶化事件+/-1年实验室确诊流感(LCI)的数据。我们用固定效应条件泊松回归估计事件的风险间隔(LCI后1-7天)与对照间隔(LCI的所有其他时间+/-1年)的发生率(IR)(95%CI)。我们纳入了用于中介分析的生物标志物数据。
结果:我们确定了3439例冠状动脉/脑血管相关住院。LCI风险与对照间隔的IRs(95%CI)为STEMI0.6(0.1,4.4),NSTEMI7.3(5.8,9.2),缺血性卒中4.0(3.0,5.4),出血性中风6.2(3.4,11.5),冠状动脉痉挛1.3(0.5,3.0)。在≥65岁的人群中,NSTEMI和缺血性卒中的IR显着增加。NSTEMI和缺血性卒中的IR下降了26%和10%,分别,当考虑白细胞(WBC)和血小板计数时。LCI与先前存在的哮喘的恶化显著相关,慢性阻塞性肺疾病,充血性心力衰竭.
结论:我们发现LCI与NSTEMI住院之间存在显著关联,缺血性卒中,出血性中风,后者可能是由于SCCS设计中下落不明的时变混杂所致。