台灣高血壓病患潛在藥品交互作用之盛行率及臨床重要性

Abstract

背景 高血壓病患常因高齡、盛行率高、降血壓藥經由肝代謝及同時有多種疾病需服藥，比一般人更易造成潛在藥品交互作用的危險。當兩種藥品發生一級潛在藥品交互作用，臨床上不一定絕對禁用，有時因病情需要併用，因此本研究將進一步，將一級潛在藥品交互作用區分為需避免併用及需調整劑量兩類，做為臨床應用的參考。另外，本研究將估計台灣高血壓病患發生潛在藥品交互作用的盛行率及常見潛在交互作用配對導致的住院率。 方法 我們採用橫斷研究估計潛在藥品交互作用的盛行率，並採用回溯性世代研究估計因藥品交互作用導致的住院。資料來源為1997年至2001年全民健保承保歸人抽樣資料庫中，高血壓病患的所有處方用藥。潛在藥品交互作用的分析範圍，包含同一張處方及不同處方(同一家醫院及不同醫院)。 本研究採用邏輯式分析，估計潛在藥品交互作用是否存在與住院發生的相對危險性。另採用chi-square分析年齡、性別、慢性病數目與用藥品項數等變項與潛在藥品交互作用的相關性。所有資料使用SAS 9.1版軟體分析(SAS Institute, Cary, NC)，潛在藥品交互作用的配對採用SAS-SQL程式。所有檢定採用雙尾檢定(α = 0.05)。 結果 資料庫中共有23,925位高血壓病患，無論一般處方箋或高血壓處方箋，以處方箋計算，潛在藥品交互作用都有逐年下降的趨勢，平均盛行率分別為10.5%及30.5%。最常見需避免併用的一級潛在藥品交互作用的藥品為cisapride，其中以不同醫院的處方發生交互作用的情形最嚴重。最常見需調整劑量的一級潛在藥品交互作用的藥品為digoxin。潛在藥品交互作用的比例與年齡、用藥品項數及合併的慢性病數目成正比。最常見與二級潛在藥品交互作用相關的藥品為降血糖的sulfonamide urea derivatives，導致住院的相對危險性為1.71 (95% 信賴區間為1.01-2.89). 結論 高齡及合併多種疾病的高血壓病患，更容易因潛在藥品交互作用，導致不良反應的危險性。本研究發現，不同醫院的處方箋比同一張處方箋的藥品，更容易因交互作用發生嚴重的不良反應而住院。若能將本研究發現的臨床上較重要的交互作用配對，建置於醫師的開方畫面，進一步若能串聯不同醫院的藥歷，避免不同處方發生交互作用的危險，將更能提升台灣病患用藥的安全性。

Background The coincidence of patients with the more elderly, the high prevalence, combination drug therapy, hepatic metabolisms of antihypertensive drugs and drug-treated comorbid conditions might result in higher possibility of many drug-drug interactions (DDIs) in patients with hypertension. Our study purposes were to estimate prevalence, incidence of hospital admission associated with common potential DDIs and identify the pairings of significance ‘1’ potential DDIs those should be avoided or need dose adjustment for drug safety. Methods We used cross-sectional study to estimate the prevalence of potential DDIs and used the retrospective cohort study to estimate the incidence of hospital admission. The data resources were all prescriptions taken by patients with hypertension in the cohort-sampling databases of national health insurance from 1997 through 2001. The potential DDIs induced from the same prescription and different prescriptions were analyzed in this study. The Logistic regression analysis was used to estimate relative risks of admission incidences. The χP2 Pstatistic was used to evaluate the association of the variables (e.g.,age, sex, number of chronic diseases and of medications). All data were analyzed with SAS 9.1 software (SAS Institute, Cary, NC) and the potential DDIs were selected with SAS-SQL program. All significant test was two-tailed with α = 0.05. Results We found 23,925 patients with hypertension in the database. The prevalences of potential DDIs by prescription in general prescriptions and hypertensive prescriptions were reduced yearly and the mean prevalence with 10.5% and 30.5%, respectively. The most common significance ‘1’ potential DDIs those should be avoided was cisapride, especially in the prescriptions from different hospital. The most common significance ‘1’ potential DDIs those need dose adjustment was associated with digoxin. The ratios of potential DDIs were proportional to age, drug items of prescriptions and concomitant numbers of chronic diseases. The most common medications associated with significance ‘2’ potential DDIs were sulfonamide urea derivatives and the admission incidence was 1.71 (95% CI 1.01-2.89). Conclusions Patients with hypertension, especially elderly and multiple illness, had higher risk to occur serious adverse drug reaction from potential DDIs. We found the most important source of potential DDIs was the prescriptions from different hospitals. The electronic file of common potential DDIs that we set up could be implemented to the computer system and integrated with the drug history from the network of different hospitals for drug safety.