使用全民健康保險研究資料庫研究warfarin處方型態 及潛在一級交互作用之風險分析

Prescribing Pattern and Risk Analysis of Potential Significant Drug Interactions among Warfarin Users in Taiwan: A National Health Insurance Research Database Study

陳雅婷(Ya-Ting Chen)

指導教授 ： 林慧玲 ； 林淑文

共同指導教授 ： 王榮德

國立臺灣大學/醫學院/臨床藥學研究所/碩士(2010年)

https://doi.org/10.6342/NTU.2010.00483

摘要

研究背景：
Warfarin之治療濃度範圍狹窄，並具有高蛋白質結合率與由cytochrom P450酵素代謝之特性，故使得使用warfarin時容易發生嚴重且危及生命之交互作用以及嚴重藥品不良事件，增加1.2∼4.5倍發生嚴重出血之風險、造成住院、或增加住院天數。然而，目前亞洲地區國家關於warfarin之研究極少，且現有國外研究之研究族群多未具有全國代表性，交互作用藥品之定義也不一致。

研究目的：
以具有全國代表性之臺灣全民健保資料庫進行分析，且考慮交互作用之臨床顯著性等級，分析臺灣地區使用warfarin之處方型態、一級交互作用之盛行率、嚴重不良事件（包括嚴重出血及血栓栓塞）之發生率。

研究方法：
使用1997∼2004年臺灣100萬人健保資料庫承保抽樣歸人檔，進行回溯性世代研究，研究族群為1997年3月31日以後開始服用warfarin的新使用者（new-users）。Warfarin處方型態分析包含每日使用劑量、平均PT/INR (prothrombin time/international normalized ratio)值監測頻率等。
藥品交互作用的發生定義為藥品的使用與warfarin治療事件期間有任何重疊者。一級藥品交互作用定義為嚴重程度屬嚴重（major）、文獻等級屬已建立的（established），根據此定義列出56種一級藥品交互作用納入分析。並分析一級交互作用發生時的處置情況，比較warfarin劑量、PT/INR監測頻率是否有差異。
嚴重藥品不良事件定義為第一次因出血或血栓栓塞造成急診就醫且（或）住院之事件，以國際疾病分類號第九版（ICD9-CM）診斷碼定義之。為釐清血栓栓塞事件是否確實為新發生的事件，本研究進一步以病人使用warfarin之適應症、急診住院是否使用抗血栓藥品校正。

研究結果：
在4047位warfarin新使用者中，平均年齡為63.4歲，女性比例稍高（56.3%），最常見之使用適應症為缺血性心臟病（23.2%），最常見之共病症為高血壓（36.4%）。平均每人每日劑量為2.8（±1.2）mg，有近三成（29.1%）的病人未在warfarin使用期間監測PT/INR，在曾接受PT/INR檢驗者中，平均每人之PT/INR監測頻率約為每40.5天監測一次。
有六成以上（63.8%）的warfarin使用者曾暴露於一級交互作用之下。前五名常見之一級交互作用藥品為aspirin (34.5%)、noscapine (11.1%)、amiodarone (9.4%)、ketoprofen (7.4%)、和naproxen (7.2%)，且近七成（67.7%）是來自同位醫師，約一半（54.4%）是來自醫學中心。並且，有無一級交互作用時的warfarin劑量無統計上顯著差異（p=0.114），且有近五成（44.6%）的病人在交互作用期間無監測PT/INR值，且當交互作用來自不同醫師時，有更高比例（66.7%）無監測。
有412位新使用者發生第一次出血事件，發生率為每人年10.9%，以腸胃道出血（40.8%）最為常見。有474位新使用者發生第一次血栓栓塞事件，發生率為每人年12.8%，以缺血性中風（31.4%）最為常見。

結論：
Warfarin使用者平均每人每日劑量為2.8 mg，且近三成無監測PT/INR，平均PT/INR監測頻率大於四週。Warfarin之一級交互作用之盛行率有六成以上，常見藥品為aspirin、noscapine、amiodarone，近五成的病人在交互作用期間無監測PT/INR值。而且，出血或血栓栓塞事件之發生率各為每人年10.9%、12.8%，最常見各為腸胃道出血與缺血性中風。由於臺灣目前尚無法全面讀取病人來自不同醫師的處方，且各院內鮮有交互作用警示系統，故極有必要建立一系統去偵測院內及跨院交互作用，提供醫療人員提示畫面及處理建議；同時進行醫師藥師之教育訓練，增進醫療品質與用藥安全。

關鍵字

warfarin ； 處方型態 ； 藥品交互作用 ； 藥品不良事件 ； 全民健康保險研究資料庫

並列摘要

Background:
Warfarin exhibits narrow therapeutic range and high protein binding. It is metabolized via cytochrom P450 enzyme system. Therefore, patients treated with warfarin are prone to serious drug-drug interactions (DDIs) and major adverse events (AEs). DDI of warfarin is significantly associated with 1.2~4.5 times increased risk of major hemorrhage, hospital admission, and prolonged length of stay. However, the occurrence of DDIs and AEs during warfarin therapy has not been studied extensively among Asian population. Few studies used national database, and definitions and lists of DDIs varied in the literature.

Objectives:
Using Taiwan’s National Health Insurance Research Database to analyze the prescription pattern of warfarin, prevalence rate of potential significant DDIs, incidence rate (IR) of major AEs (i.e., hemorrhage and thromboembolism) among warfarin users in Taiwan.

Methods:
The National Health Insurance Research Database with one million randomly selected individuals in Taiwan was used in this retrospective cohort study. New users started warfarin between April 1997 and December 2004. Prescribing pattern, including prescribed daily warfarin dose and average PT/INR (prothrombin time/international normalized ratio) monitoring frequency, were analyzed.
DDIs were defined as prescriptions from outpatient clinics with at least one day overlap with warfarin treatment episodes. Significance grade one DDIs were defined as DDIs with major severity and established documentation. A total of 56 grade one DDIs were evaluated in this study. In addition, dose of warfarin and PT/INR monitoring frequency when DDIs occurred were compared with those without grade one DDIs.
Major adverse events were defined as the first emergency room visits and/or hospitalizations with a diagnosis of hemorrhage or thromboembolism, identified by International Classification of Disease, Ninth Revision, Clinical Modification (ICD9-CM), during the treatment episode. Moreover, all thromboembolism events were confirmed by antithrombotic agents and primary indications for warfarin.

Results:
Among 4047 new-users, mean age was 63.4 years old. More than half (56.3%) of new-users were female. Warfarin was prescribed mainly for ischemic heart disease (23.2%). The most common comobidity was hypertension (36.4%). Average daily dose of warfarin was 2.8 ±1.2 mg. There were 29.1% of patients never received PT/INR monitoring during warfarin treatment. Other patients monitored PT/INR every 40.5 days.
A total of 2582 warfarin users (63.8%) were concurrently prescribed drugs with grade 1 DDIs. The most frequently prescribed medicines were aspirin (34.5%), noscapine (11.1%), amiodarone (9.4%), ketoprofen (7.4%), and naproxen (7.2%). Approximately 70% of these prescriptions were from the same physicians. About half of DDIs (54.4%) were from the medical centers. Doses of warfarin with or without DDIs were similar (p = 0.114). About half of patients (44.6%) had no PT/INR monitoring orders during DDIs. Fewer patients checked PT/INR when DDIs were involved different physicians.
There were 412 patients developing hemorrhage with an incidence rate of 10.9% per person-year, and the majority was gastrointestinal hemorrhage (40.8%). On the other hand, 474 patients experiencing thromboembolism with an incidence rate of 12.8% per person-year, and the most common diagnosis was ischemic stroke (31.4%).

Conclusions:
Average daily warfarin dose was 2.8 mg. Approximately 30% of patients had no PT/INR monitoring during treatment. PT/INR was ordered about every 6 weeks. Prevalence of grade one DDIs were more than 60%. The most prevalent DDIs were aspirin, noscapine, amiodarone. Almost half of patients had no PT/INR monitoring orders during DDIs. Moreover, IRs of hemorrhage and thromboembolism were 10.9%, 12.8% per person-years, respectively. The most common major AEs were gastrointestinal hemorrhage and ischemic stroke. In Taiwan, medical records from different physicians have not yet been completely retrieved through medical insurance cards. Computerized DDI alert systems are uncommon in hospitals. It is essential to establish a computerized electronic alert system on DDIs along with continuing educations for physician and pharmacist to safeguard patients’ safety.

並列關鍵字

warfarin ； prescribing pattern ； drug interaction ； drug adverse event ； National Health Insurance Research Database

參考文獻

43. Yu HY, Liu CH, Chen YS, Wang SS, Chu SH, Lin FY. Relationship of international normalized ratio to bleeding and thromboembolism rates in Taiwanese patients receiving vitamin K antagonist after mechanical valve replacement. J Formos Med Assoc 2005;104:236-43.

117. Yen YH. Establishing a computerized protecting network of drug safety. Journal of Healthcare Management 2004;5:495-514.

1. Ansell J, Hirsh J, Hylek E, Jacobson A, Crowther M, Palareti G. Pharmacology and management of the vitamin K antagonists: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest 2008;133:160S-98S.

2. Wysowski DK, Nourjah P, Swartz L. Bleeding complications with warfarin use: a prevalent adverse effect resulting in regulatory action. Arch Intern Med 2007;167:1414-9.

3. Harder S, Thurmann P. Clinically important drug interactions with anticoagulants. An update. Clin Pharmacokinet 1996;30:416-44.