| 摘要: | 研究目的︰1) 針對容易因交互作用而產生不良反應的高風險藥物，包括抗凝血藥物、抗血小板藥物及digoxin等，建立中西藥交互作用之實證資料庫；2) 依其結果探討2005年期間國人中西藥併用及發生潛在交互作用的情形及相關因子，再針對抗血小板藥物使用者，探討有無併用西洋參、人參、丹參或當歸等中藥以及與腸胃道出血發生之相關性。 研究方法︰首先利用系統回顧方法，經結構性文獻搜尋步驟，自一級至三級文獻資料庫及相關網站中摘錄相關之實證，經彙整、評估及專家審查後建立實證資料庫。接著利用健保資料庫進行族群回顧性世代研究，針對2005年曾於健保給付的處方或被調劑過有抗凝血藥物 (heparin, warfarin)、抗血小板藥物 (aspirin, clopidogrel, dipyridamole, ticlopidine) 或 digoxin 之門診病人，利用描述性統計分析中西藥併用和潛在中西藥交互作用發生之盛行率和使用型態，以卡方檢定 (χ2 test) 和 t 檢定 (t test) 比較病患特性、藥品使用品項、醫療花費、醫療照護使用情形之差異，並利用羅吉斯迴歸評估影響的相關因子。最後則是利用嵌入型病例對照研究，在67,401位使用抗血小板藥物且無腸胃道出血病史的病患中，最後針對559位曾在 2005 年因腸胃道出血而住院的病人，以一比一配對方式篩選對照組後，利用條件式邏輯斯迴歸來評估其腸胃道出血的危險，與併用有潛在交互作用之中藥是否有相關性。 結果︰自8本書籍、187篇雜誌文章或論文及5個網站中，共獲得 228 對個別高風險藥物與中藥之交互作用；其中有 122 對 (53.5%) 可在相關資料庫查到機轉和嚴重度的資訊，有高達 109 對 (89.3%) 的交互作用是屬於藥物藥效學的機轉。2005年的高風險藥物使用者有 70,698，併用中藥的盛行率為13.22% (9,349人)，發生率為6.33%；會增加校正後的高風險藥物併用中藥的勝算比的因子包括女性、45-55歲、中等程度的經濟狀況、在中區分局加保、每個月看過2次以上門診、曾使用超過20種藥物及曾診斷為中風和高血壓者。在9,349個曾在2005年併用高風險藥物與中藥的病人中，有潛在重要之中西藥交互作用存在者有7,566人 (80.93%)；會增加校正後的潛在中西藥交互作用發生勝算比的因子包括45-84歲、每個月看過3次以上門診、曾使用20-39種或60-79種藥物及曾診斷為關節炎者。反之，於中區分局加保者或曾診斷為急性呼吸道感染者發生潛在中西藥交互作用的危險則較低。最後，抗血小板藥物與中藥潛在交互作用對腸胃道出血的危險性分析結果顯示，抗血小板藥物與人參或當歸併用時，有增加10%~74%腸胃道出血的危險性，但因為樣本數較少並無統計上顯著的差異。 討論與建議︰本研究所建立的實證資料庫發現，抗凝血藥物、抗血小板藥物或 digoxin等高風險藥物與許多中藥間存在著具實證依據的潛在交互作用。經系統回顧已確認有122 對交互作用被歸類為嚴重或中度程度。其中抗凝血藥物或抗血小板藥物與中藥交互作用的結果大多是增加出血的危險；而 digoxin 與中藥交互作用的結果亦多是加強其作用。這樣的中西藥交互作用大都歸類為藥物藥效學反應，因此即使錯開使用仍有可能還是會發生不良反應。本中西藥實證資料庫的建立，是為了要提供給臨床工作者一個摘要，尤其是容易發生潛在交互作用的一些高風險中西藥品項；但是醫事人員在判斷臨床重要性時，除參考資料庫的資訊外，建議應該依據病人的實際情形及臨床專家的經驗加以評估。在高風險藥物使用者中，有13% 病人曾與中藥併用，其中又有81%暴露於潛在的中西藥交互作用中。抗血小板藥物與人參或當歸併用時，有增加10%~74%腸胃道出血的危險性，但因為樣本數較少且並無統計上顯著的差異。本研究探討臺灣地區病人併用高風險藥物與中藥、及發生潛在交互作用之真實面貌，除可提供具體的統計數據，讓國人對此議題有更深切地體認外；所發現的相關因子更可提供作為醫療照護人員及民眾/病患間溝通的參考依據。本研究所建立之先驅模式亦可作為下一階段研究之依據，針對更長的研究期間來研究。在此更建議醫療照護者應特別留意這些病患的臨床反應，並應加強與病患的溝通與衛教。
Aim: 1) To examine and evaluate the documented evidence of potential interactions between high risk western medications (HRWM, i.e., aspirin, clopidogrel, digoxin, dipyridamole, heparin, ticlopidine, and warfarin) and Chinese Medicines (CM); 2) to examine the use patterns of HRWM with concentrated CM and the exposure of potential interactions using National Health Insurance Research Database (NHIRD) databases to explore associated factors; and to determine the gastrointestinal (GI)bleeding risks due to exposure to the concurrent use of antiplatelet drugs with CM (i.e., American ginseng, Asian ginseng, danshen, and dong quai). Method: We conducted a structured literature search on tertiary literature, primary literature, and the relevant web resources. We extracted all relevant data regarding the harmful HRWM-CM interactions using the standardized checklist. Possible mechanisms and severity ratings of documented interactions were extracted using MicroMedex, Lexi-Interact?, and Natural Medicines Comprehensive Database. A cohort study using the 2005 NHIRD in Taiwan was conducted. Concurrent use of concentrated CM, exposure to potential HRWM-CM interactions, demographics, comorbidities, and health service utilization among HRWM users were evaluated. The characteristics were compared using t tests and Chi-square tests. Multivariate logistic regression was performed to explore factors associated with concomitant HRWM-CM use and exposure to potential interactions. Then, a nested case-control study was conducted in a cohort of antiplatelet drugs users without history of GI bleeding. Cases were patients who were hospitalized due to GI bleeding in 2005. Conditional logistic regression analysis was used to determine the association between GI bleeding and exposure to interactions between antiplatelet agents and specific CM. Results: Of the retrieved eight books, 187 primary literature and five web sites, a total of 228 unique pairs of HRWM-CM interactions were reviewed. Of 122 HRWM-CM interaction pairs being described its mechanism and severity, 89.3% were due to altered pharmacodynamics. Of the 70,698 HRWM users, 13.22% used CM currently. After excluding those who had a prior CM use, the incidence of concurrent use was 6.33%. Factors that increased the odds of concurrent HRWM-CM use included female sex, age 45-54 years, middle socio-economic status, enrollment in Central Taiwan, higher number of outpatient visits or distinct medications during the baseline period, and previous diagnoses of stroke and hypertension. In contrast, age ?65 years and medical expenditure ?50,000NTD were associated with reduced odds of concurrent HRWM-CM use. Among 9,349 HRWM-CM concurrent users, 80.93% (7,566) had ever been exposed to potential HRWM-CM interactions. Factors that increased the odds of exposure to potential interactions included age 45-84 years, higher number of outpatient visits or distinct medications during the baseline period, and previous diagnoses of arthritis. In contrast, enrollment in Central Taiwan, and previous diagnosis of acute respiratory infections were associated with reduced odds. Finally, the antiplatelet drugs users were 10% to 74% more risk of GI bleeding among those who had used Asian ginseng, or dong quai currently. However, there were no statistically significant differences after adjustment for these uses. Discussion: We found that a bunch of evidence showed the interactions between anticoagulants, antiplatelet drugs, or digoxin and several CM. Of 122 pairs of moderate to severe HRWM-CM interactions, the majority of interactions were pharmacodynamic effects and could not be avoided even if the HRWM and CM were taken in the different times. These findings would provide the practitioners with a summary to identify the risky combinations. However, clinicians should integrate their individual expertise to differentiate the similarities and difference between the recommendations in this review paper and other evidence. In Taiwan, 13% of HRWM users were prescribed with CM concurrently, and 81% of the concurrent users were exposed to potential HRWM-CM interactions. Whether or not such concomitant use is associated with adverse clinical outcomes warrants further investigation. This is one of the studies to examine the actual use patterns of HRWM and CM on the population base in Taiwan, the findings will be used for further or next phase research. |
