目的：本研究目的是藉由分析與探討藥物過敏潛在危險因子，提供臨床評估參 考，以降低藥物過敏的風險。 方法：透過藥物不良反應通報系統，取得自2012 年 1 月 1 日至2016 年 12 月 31 日期間，納入疑似藥物過敏的個案，採用回溯性病歷研究。類別變項以卡方檢定 (chi-square test) 分析，並以多元邏輯斯迴歸 (multinomial logistic regression) 分析危 險因子與嚴重度之相關性。 結果：本研究共納入 1,019 人，平均年齡為 53.7 ± 21.5 歲；其中男性占 51.4%。 在基本人口學變項與醫療資源之分布分析，不論在年齡、用藥天數、給藥途徑、就醫 型態和疑似過敏藥物，皆與過敏嚴重程度有顯著差異 (p = 0.003, < 0.001, < 0.001, < 0.001, and < 0.001)。多變項分析中，在重度與輕度組比較，發生在用藥天數 > 1 天 有顯著差異 (odds ratio [OR] = 2.94; 95% confidence interval [CI] = 1.280 ~ 6.750)。 門診病人與住院病人過敏症狀不論中度與輕度 (OR = 0.45; 95% CI = 0.330 ~ 0.620) 或重度與輕度 (OR = 0.01; 95% CI = 0.001 ~ 0.063) 組比較，皆較輕微。疑似藥物分 類以抗生素 (39.5%) 和中樞神經用藥 (34.2%) 為主。藥物過敏臨床症狀以皮膚表現為 主 (72.5%)。 結論：本研究經由分析的藥物過敏危險因子，其中年齡、給藥途徑、用藥天數 及個別藥物品項等，與文獻有共通性。期望未來收集更多關於藥物過敏的個案及對照 組，進一步分析潛在危險因子和發病率，以提供臨床決策之參考。

ABSTRACT 

Objective: The study aimed to investigate the potential risk factors of drug allergy and provide clinical assessment to reduce the risk of drug allergy. Methods: A retrospective analysis search on adverse drug reactions was conducted from January 2012 to December 2016 to identify cases of drug-induced allergy. Chisquare test was for categorical variables, and multinomial logistic regression was adopted to verify the correlation among the severity and risk factors. Results: Chart reviews for 1,019 patients were conducted. Mean age of patients was 53.7 ± 21.5 years and 51.4% were men. Regardless of age, time for onset of drug reaction, route of administration, patient type, and culprit drug classification, significant difference between the severity of drug allergy (p = 0.003, < 0.001, < 0.001, < 0.001, and < 0.001) was found. In the multivariate analysis, onset of drug reaction more than 1 day was significantly higher in the severe vs. mild group (odds ratio [OR] = 2.94; 95% confidence interval [CI] = 1.280-6.750). Meanwhile, severity of drug allergy in outpatients was slightly higher than that in inpatients. Conclusions: The present study points out the common risk factors for drug allergy including age, route of administration, time for onset of drug reaction, and culprit drug classification. To help the physician to valuate further the risk of developing drug allergy, cases of drug allergy with control group studies shall be needed.