2011-08-012024-05-14https://scholars.lib.ntu.edu.tw/handle/123456789/658931摘要：美國每年有四百萬名病患感染社區肺炎，社區肺炎具有高罹病率、高死亡率，造成了高醫療與社會成本。肺炎是急診醫學領域最重要的疾病之一，高達百分之七十五的社區肺炎在急診處接受第一線治療。儘管如此，過去對於社區肺炎的危險因子研究卻相當有限，除了已知的酗酒、免疫不全、心肺疾病外並沒有新發現可避免的危險因子。然而最近的研究顯示，使用氫離子幫浦抑制劑可能是社區肺炎的危險因子之一。憑藉著強效性與優良的藥物動力學，氫離子幫浦抑制劑在上市後很快就取代乙型組織胺阻斷劑成為最普遍的制酸劑，但是過去十五年間也一直存在著長期使用的安全性疑慮，目前已經證實長期使用氫離子幫浦抑制劑會導致感染性大腸炎，同樣對於氫離子幫浦抑制劑導致社區肺炎的疑慮，相關研究卻無法得到一致結論。由於藥物的長期併發症並無法運用臨床隨機試驗證實，所以觀察性研究成了藥物安全研究的最重要工具。觀察性研究困難之處在於必須控制干擾因子方能得到真實不偏的相關預估值。近年來氫離子幫浦抑制劑有過度使用的趨勢，不僅是潰瘍性疾病許多情況下，例如壓力性潰瘍的預防，或是非類固醇性抗發炎藥物的潰瘍預防都會給予氫離子幫浦抑制劑處方；另一方面，肺炎的病因異質化也是很棘手的問題。肺炎是許多慢性病例如中風、癌症、糖尿病等的共同併發症。可以預見氫離子幫浦抑制劑使用和肺炎之間存在許多共同干擾因子，增添觀察性研究預估值的不穩定性。為了要釐清氫離子幫浦抑制劑使用和肺部感染的關係，我們認為探討病因專屬的肺部感染而非廣泛性的肺炎做為研究目標，應是觀察性研究較合理的設計。氫離子幫浦抑制劑使用和結核病的可能相關，還可以得到生物學上的支持。1960 年代胃切除是治療胃潰瘍普遍的方法，但是外科醫師卻發現經胃切除的病患在後續六個月內確有高比例產生結核病。從那時起胃切除便成了廣為人知的結核病危險因子，雖然致病機轉尚未釐清但是一般認為和胃酸對結核菌的抑制應該有相當的關聯。臨床上培養結核菌的標本來源除了痰液外便屬胃液的培養率最高。結核病是開發中國家負擔最大的疾病，台灣目前為止也還屬於高感染區。本研究的結果對於公共衛生當有相當的啟發性。我們第一階段將進行統合分析(Meta-analysis)，將目前所有的觀察性研究做一歸納分析探討結論異質化的原因。其次我們將運用台灣健保研究資料庫進行全國代表性樣本的研究，我們將運用哈佛大學藥物流行病學教授S. Sebastian 的新方法—高維度傾向計分析對干擾因子做全面的控制，由於此方法學上的創新才剛發表相關應用的研究並非很多。所以我們另一方面也會運用傳統巢式病例對照研究分析，提供後續學者研究一個比較的基礎。<br> Abstract: Community-acquired pneumonia causes more than 4 million episodes of illnesseach year and has high morbidity, mortality, and total cost of care. Nationwide, nearly75% of community-acquired pneumonia patients are initially evaluated and treated inhospital-based emergency departments (EDs). Population-based studies of risk factorsfor CAP are scarce. Except for the know risk factors such as alcoholism, heart disease,lung disease, and immunosuppressive therapy, use of gastric suppressant, especiallyproton pump inhibitors (PPIs), were recently identifies as an potential risk factors forpneumonia.Due to high potency and advantageous pharmacokinetic profile, PPIs have replacedhistamine receptor antagonists as the mainstay of treatment for and prevention of manyserious gastrointestinal diseases. However, there has been continued concern regardingthe long-term safety of PPIs. Increased susceptibility of infections has been reported aspotential complications, which is theorized to be the result of allowing survival ofpotentially pathogenic organisms due to lowered gastric pH. Recent observationalstudies, however, did not show consistent findings of the association between gastricacid suppression and community-acquired pneumonia. One of major methodologicalchallenge is confounding by indication. PPIs are overused in clinical practice, whichmakes them a proxy of general health rather than a proxy of ulcer-related disease.Therefore, the use of PPIs shared a lot of common caused with pneumonia and caused alot of unobserved confounding.To clarify the relationship between gastric-acid suppression and respiratory tractinfection in an observational studies, we believe, a microbiologically based diagnosis ofpulmonary infection is necessary, as has been demonstrated by the studies on PPI useand enteric infections. Hypothetical relationship between PPI use and active TB maylend support from the clinical observations that a significant number of gastrectomypatients developed active TB subsequently. Although the reasons for the associationhave not been clarified, it is possible that gastric acid may have played an important rolein the suppression of TB reactivation because patients with active TB have a highpositive culture yield of M. tuberculosis from the gastric lavage fluid.TB is today considered as the most important health problem especially indeveloping countries. It is estimated that one third of the world population is infectedwith Mycobacterium tuberculosis. Despite the greet advance of public health in recentyears, Taiwan remains one of the TB endemic area. Therefore, the aim of our study mayhave great public health and medical implications. In this proposal, we seek to addressthe research question by first performing a meta-analysis summarizing all availableobservational studies on the association between PPI use and pneumonia. Then we willperform a population-based study with the National Health Insurance ResearchDatabase (NHIRD) of Taiwan. We would use high dimensional propensity score (HDPS)developed by Schneeweiss S. et al to use the rich information of a health claim databasefor controlling of confounding. Then the results would further compared with a new usercohort study.氫離子幫浦抑制劑組織胺-2接受器拮抗劑肺炎肺結核proton pump inhibitorhistamine receptor 2 antagonistspneumoniatuberculosis