使用可能造成低凝血酶原血症的Cephalosporin類抗生素與出血風險：全國性嵌入型病例對照研究

Abstract

研究背景： 過去有許多文獻顯示，帶有N-methylthiotetrazole（NMTT）、methylthiadiazolethiol（MTDT）及N-hydroxyethyltetrazolethiol（HTT）等支鏈的cephalosporin類抗生素可能造成低凝血酶原血症（hypoprothrombinemia），而導致發生凝血功能異常的副作用。然而，目前的研究侷限於病例報告及系列病例報告，近年來亦缺乏以族群為基礎的研究。 研究目的： 本研究之目的係希望透過全民健康保險資料庫，研究使用cefmetazole、cefotetan、cefoxitin、flomoxef、moxalactam、cefamandole及cefoperazone等過去文獻曾報告過可能增加出血傾向或發生出血事件副作用之cephasporin類抗生素的病人群，評估其發生出血或增加出血傾向副作用之比例，以及該副作用與使用此類抗生素之相關性；並探討使用此類抗生素而發生出血或增加出血傾向之危險因子。 研究方法： 本研究期間設定為2000年1月1日至2011年12月31日，方式為嵌入型病例對照研究。研究之納入條件為在研究期間內年齡20歲以上、在急診使用針劑劑型的cefmetazole、cefotetan、cefoxitin、flomoxef、moxalactam、cefamandole、cefoperazone、amoxicillin/clavulanate、ampicillin/sulbactam、cefuroxime、cefotaxime和ceftriaxone等抗生素48小時以上，且在健保資料庫內有過就醫紀錄之病人。 本研究定義於研究期間內，有發生病例組定義者，包括：住院期間以International Classification of Diseases, Ninth Revision, Clinical Modification code（ICD-9-CM code）找出有出血診斷者，或以全民健保藥品給付代碼（NHI code）找出使用維生素K1、第二、第七、第九和第十凝血因子或新鮮冷凍血漿（fresh frozen plasma, FFP）者為病例組；而於研究期間內，沒有發生病例組定義者則為對照組，並同時定義病例組與對照組該次住院日期為index date。本研究將病例組與對照組依年齡、性別及index date進行1:4配對，而使用特定cephalosporin類藥品與出血傾向增加或發生出血事件的相關性，則透過conditional logistic regression進行分析。 研究結果： 本研究之研究族群共有6,191人，其中，經配對後得到的病例組共704人，而對照組則有2,816人。研究結果顯示，使用可能造成低凝血酶原血症的cephalosporin類抗生素與出血傾向增加或發生出血事件具統計上顯著相關（adjusted OR 1.71; 95 % CI [1.42-2.06]），且這類cephalosporin抗生素中，使用cefoperazone（adjusted OR 4.57; 95 % CI [2.63-7.95]）、cefmetazole（adjusted OR 2.88; 95 % CI [2.08-4.00]）及flomoxef（adjusted OR 1.35; 95 % CI [1.09-1.67]）的風險最高。再進一步將累積劑量分為 < 3 DDD、≥ 3 DDD且< 5 DDD及≥ 5 DDD三組，可看出增加出血傾向和發生出血事件的風險隨著累積劑量增加而增加（adjusted OR 1.62; 95 % CI [1.28-2.05] versus adjusted OR 1.78; 95 % CI [1.35-2.34] versus adjusted OR 1.89; 95 % CI [1.28-2.78]）。另外，病人於index date前6個月內有以下2種情況發生：(1) 出血事件（adjusted OR 2.57; 95 % CI [1.94-3.41]）；(2) 本身營養狀況較差（adjusted OR 1.41; 95 % CI [1.15-1.73]）；於index date前6個月有使用以下2種藥品：(1) 抗凝血劑（adjusted OR 2.08; 95 % CI [1.64-2.63]）；(2) 肝衰竭用藥（adjusted OR 1.69; 95 % CI [1.30-2.18]）等4種情形，為此類藥品副作用發生之風險因子。 結論： 由於使用cefmetazole、cefotetan、cefoxitin、flomoxef、moxalactam、cefamandole及cefoperazone等抗生素相較於使用amoxicillin/clavulanate、ampicillin/sulbactam、cefuroxime、cefotaxime和ceftriaxone等抗生素顯著增加出血傾向提高或發生出血事件的風險，因此需謹慎處方這類抗生素，尤其當病人有出血事件史、營養狀況不良、過去6個月內曾服用抗凝血劑、或有肝功能衰竭病史時，在處方這類抗生素時，應定期監測病人的國際標準化凝血酶原時間比值（international normalized ratio，INR）。

Background: The results of previous studies have shown N-methylthiotetrazole (NMTT)-side-chain-, methylthiadiazolethiol (MTDT)-side-chain-, and N-hydroxyethyltetrazolethiol (HTT)-side-chain-containing cephalosporins caused hypoprothrombinemia and then led to coagulopathy. However, most of the previous studies are limited to case reports and case series studies, and no population-based study was to examine such association. Aims: The aims of this study was to investigate the association between the use of intravenous hypoprothrombinemia-inducing cephalosporins and the risk of hemorrhagic tendency or events, and to identify possible risk factors using the National Health Insurance Research Database (NHIRD) in Taiwan. Methods: We identified adult patients receiving cefmetazole, cefotetan, cefoxitin, flomoxef, moxalactam, cefamandole, cefoperazone, which were defined as study antibiotics, and intravenous amoxicillin/clavulanate, ampicillin/sulbactam, cefuroxime, cefotaxime, ceftriaxone, which were identified as reference antibiotics, for more than 48 hours in the emergency department between the year 2000 and 2011 from NHIRD. Cases were defined as patients with hemorrhagic tendency or events during hospitalization subsequent to use of the abovementioned antibiotics, and matched controls were selected with the ratio of 1:4 by gender, age, and index date. The association between the exposure to study antibiotics and the risk of hemorrhagic tendency (using vitamin K, coagulation factor II, VII, XI, X, or fresh frozen plasma) or events identified through national health insurance codes (NHI codes) and International Classification of Diseases, Ninth Revision, Clinical Modification codes (ICD-9-CM codes) was evaluated using conditional logistic regression models. Results: Among 6191 patients in the cohort, 704 cases and 2816 matched controls were identified. The conditional logistic regression model showed the risk of hemorrhagic tendency or events was associated with exposure to study antibiotics (adjusted OR 1.71; 95 % CI [1.42-2.06]), and cefoperazone (adjusted OR 4.57; 95 % CI [2.63-7.95]), cefmetazole (adjusted OR 2.88; 95 % CI [2.08-4.00]) and flomoxef (adjusted OR 1.35; 95 % CI [1.09-1.67]) were at the highest risk among all study antibiotics. In addition, after classifying cumulative dose of study antibiotics into 3 groups, which were less than 3 DDD, more than 3 but less than 5 DDD, and more than 5 DDD, the result revealed that patients using higher cumulative dose of study antibiotics had higher risk of hemorrhagic tendency or events (adjusted OR 1.62; 95 % CI [1.28-2.05] versus adjusted OR 1.78; 95 % CI [1.35-2.34] versus adjusted OR 1.89; 95 % CI [1.28-2.78]). Moreover, use of anticoagulants (adjusted OR 2.08; 95 % CI [1.64-2.63]), liver failure (adjusted OR 1.69; 95 % CI [1.30-2.18]), poor nutritional status (adjusted OR 1.41; 95 % CI [1.15-1.73]), and history of hemorrhagic events (adjusted OR 2.57; 95 % CI [1.94-3.41]) within 6 months prior to the index date were significant risk factors of the side effect brought by study antibiotics. Conclusions: Study antibiotics, including cefmetazole, cefoxitin, flomoxef, cefotetan, moxalactam, cefamandole, and cefoperazone, should be prescribed prudently due to an increased risk of hemorrhagic tendency or events compared to reference antibiotics in hospitalized patients due to community-acquired infections. Close monitoring of INR levels is recommended, especially in patients who are on anticoagulants, with poor nutritional status, underlying liver failure, or history of hemorrhagic events.