- 學位論文
感染非傷寒性沙門氏桿菌患者發生深層靜脈血栓、肺栓塞和梗塞性腦中風的風險研究
The Risk of Subsequent Deep Vein Thrombosis, Pulmonary Embolism and Ischemic Stroke in Patients with Nontyphoidal Salmonellosis
摘要
血栓形成是感染症常見的伴隨結果，且與患者預後不良有關。 但是，對與感染有關的血栓形成的引發，增幅和消退的機制了解甚少，感染相關血栓形成與傳統情況下血栓形成之間的主要區別在於，病原體及其產物引起強烈的炎症。這種炎症觸發了血小板的活化，這可能伴隨對內皮的損害，導致纖維蛋白沉積和血栓形成。近期的非傷寒沙門桿菌老鼠模型顯示了此菌會引發時程較久的血栓，甚至長於數週。另外SCIENCE期刊也顯示此菌可能會潛伏在巨噬細胞內，並改變免疫機制，得以存活更久，一旦環境允許，容易復發造成疾病。我們希望以流行病學為前導，進一步探究非傷寒沙門桿菌感染在真實世界的血栓事件。 第一部分的研究: 為了評估台灣人群中診斷為非傷寒沙門桿菌病（NTS）的病患中後續發生深靜脈血栓（DVT）和肺栓塞（PE）的風險。使用住院申報數據分析了新診斷NTS的17,855例患者和沒有NTS的71,420例患者的風險。並排除在納入前有深靜脈血栓或肺栓塞病史的患者。首先兩組按照年齡，性別和住院日期進行個別匹配。應用Cox比例風險回歸模型估算DVT和PE的風險，並考慮競爭事件（死亡）。NTS組的追蹤期間為4.94（±3.93）年，非NTS組為6.30（±3.67）年，DVT和PE的調整後亞危險比（SHRs）為1.83（95％CI 1.44–2.31 ）和1.84（95％CI 1.30–2.60）。在所有年齡組中，與對照組相比，NTS組的DVT和PE風險增加。分層分析顯示，與非NTS組中相同年齡層的患者相比，NTS組中18-39歲的患者俱有更高的DVT和PE風險（DVT為aHR，5.95； 95％CI，2.22-15.91； aHR 6.72； 95％CI，PE為2.23–20.30）。進一步分層分析，對於DVT，年齡與NTS暴露之間的相互作用的P值為<0.001; 對於PE，年齡與NTS暴露之間的相互作用的P值為0.004。 此研究再以傾向分數進行配對（PSM），重新分析，兩種分析交叉驗證得到一致的結果。第一步分得研究結論: 儘管發生靜脈栓塞的絕對風險並不高，但應可提供臨床醫護有用的建議:當病患同時具有其他VTE風險因素時，有NTS住院病史的患者有更高發生VTE的風險。 第二部分的研究: 探討非傷寒性沙門桿菌病（NTS）與急性缺血性中風（IS）的風險之間的關係。使用台灣全民健康保險研究資料庫（NHIRD）其中的百萬人口抽樣檔(LHID)，進行2000年1月1日至2012年12月31日的回溯性世代研究；18歲及18歲以上有NTS感染史（N = 1,618）的患者，沒有感染紀錄的非NTS組（N = 6472）通過傾向分數進行匹配。定義追蹤期為從最初診斷NTS到發生急性缺血性中風(IS)事件，死亡或2013年12月31日為止的時間區間。排除先前診斷為中風的患者。校正人口統計學和重要的共變項後，使用帶有Cox比例風險回歸模型來估計IS的風險比（HR）和95％信賴區間（CI）。與非NTS組相比，NTS組的整體中風(含梗塞性與出血性中風)之調整後危險比（aHR）為1.34（95％置信區間[CI]，1.10-1.65）。此外與非NTS組相比，NTS世代患IS的風險增加（aHR，1.42； 95％CI，1.14-1.78）。進一步分層分析顯示，與非NTS組相比，年齡≥65歲的患者，女性，高血壓，高脂血症，COPD，癌症和心衰竭的人群發生中風的風險更高。在追蹤期的前3-12個月中觀察到中風的風險顯著增加1.93倍（95％CI，1.07-3.48）。與非NTS患者相比，NTS患者發生缺血性中風的風險增加。這個研究建議臨床醫護:對於NTS患者，提供及時而謹慎的治療計劃可能是預防中風的一個里程碑。未來仍需進一步的研究來探索潛在機轉。
關鍵字
none並列摘要
Part I: For the purpose of examination of the risk of subsequent deep vein thrombosis (DVT) and pulmonary embolism (PE) in participants with a previous hospitalization diagnosis of nontyphoid salmonellosis (NTS) in Taiwan. Relative risks were examined in one group (N=17 855) newly having nontyphoid Salmonella infection and another group (N=71 420) never with a medical diagnosis of NTS. We utilized a subset of National Health Insurance Research Database (NHIRD)- that is hospitalization claim dataset. In the first mode, two groups were retrieved by frequency-matched in terms of age, sex, and index date as the main analysis. A regression model (Cox proportional hazards) was applied to calculate the relative risk of deep vein thrombosis and pulmonary embolism, took competing risk event (death) into account. To total time of follow up time 4.94 (±3.93) years and 6.30 (±3.67) years in the study and control group, respectively. The adjusted sub-hazard ratios (SHRs) of deep vein thrombosis and pulmonary embolism were 1.83 (95% CI 1.44–2.31) and 1.84 (95% CI 1.30–2.60). The NTS patients had a higher risk of deep vein thrombosis and pulmonary embolism in comparison with the non-NTS patients; and there was no difference in subgroup analysis by age. Further stratification revealed that participants age eighteen to thirty-nine years in the NTS cohort pose statistically significant increased risk of subsequent development of deep vein thrombosis and pulmonary embolism in comparison with participants with identical age in the control group (adjusted HR 6.72; 95% CI, 2.23–20.30 for pulmonary embolism, and 5.95; 95% CI, 2.22–15.91 for deep vein thrombosis). The subgroup analysis (age) showed that the interaction effect (P-value <0.001 in deep vein thrombosis; 0.004 for pulmonary embolism). The findings were robust due to cross-validation made by another analysis (in which both cohorts were matched by propensity scores matching) from the entrance of study, and also showed positive association results. Indeed, the absolute numbers of thromboembolism are relatively low, the physicians should notice and inform people with administrative medical history of nontyphoid salmonellae infection are risky in developing venous thromboembolic events, especially whenever in the presence of traditional risk factors of thrombosis. Part II: To evaluate whether a positive association present between nontyphoid Salmonellae infection and the developing subsequent cerebrovascular event (that is stroke, including ischemic and hemorrhagic stroke; IS and HS respectively) by using Taiwanese population-based nationwide study. This cohort study which was retrospectively using Longitudinal Health Insurance Research Dataset (LHID). This is a subset derived from National of Taiwan Health Insurance Research Database (NHIRD), participants aged eighteen years and above having previous diagnosis of nontyphoidal salmonellosis (study group number 1 618) were extracted from 2000 January to 2012 December, and comparison group (non NTS number 6 472) matching both at a ratio of one to four by propensity scores. Each on study individual was matched to four comparison subjects. The time of follow-up was counted as the period from the initial medical record of NTS to the main outcome occurrence, diagnosis date of ischemic stroke, and hemorrhagic stroke, expired or 2013 December. Participants with previous medical history of stroke were not included. The regression analysis (Cox proportional hazard) along with robust-sandwich-estimator were adopted to calculate adjusted hazard ratio and 95% C.I. of events of stroke after demographic and important clinical covariates mentioned adjustment in this study. The adjusted hazard ratio in NTS group compared with the control group revealed 1.34-fold risk (95% C.I., 1.10-1.65) to get stroke. In detail, the study group held a higher risk of ischemic stroke in comparison with the control group (adjusted hazard ratio 1.42; 95% C.I., 1.14-1.78). In stratification analysis demonstrated individuals in elder group (65 years old and older), women, and patients with medical record of COPD, dyslipidemia, hypertension, malignancy and congestive heart failure are risky of subsequent new onset of cerebrovascular accident. There is a hazard ratio of 1.93 to get new onset cerebrovascular events especially at the first 3-12 months-interval after the index date. We concluded that having had a medical record of NTS shows a higher risk of subsequent cerebrovascular occlusion episodes than the control (patients without a medical history of Salmonella infection). This study suggested that timely education of hygiene food behavior could shed a light on preventing cerebrovascular accident.