懷孕婦女之藥品處方型態與相關不良事件：以安胎藥品ritodrine為例

Prescribing Pattern and Adverse Events of Drugs Among Pregnant Women: Taking the Tocolytic Drug Ritodrine as an Example

林芝琬(Chih-Wan Lin)

指導教授 ： 蕭斐元

國立臺灣大學/醫學院/臨床藥學研究所/博士(2021年)

https://doi.org/10.6342/NTU202104165

摘要

研究背景：過去的研究顯示，有相當比例的婦女會在懷孕期間使用藥物，且藥物使用比例有隨年代上升的趨勢。而與一般病人族群相比，藥品使用於懷孕婦女之效益與風險平衡需要更審慎全面的評估，除了需考量藥物可能通過胎盤進入胎兒體內，導致流產或先天性缺陷等永久性傷害，藥物不良事件也可能危害孕婦本身的健康，此外，罹患慢性疾病的婦女若因懷孕而貿然停止用藥，反而可能對母親或胎兒產生不利影響。然而，針對懷孕婦女之用藥安全資訊卻相當缺乏，主要原因來自動物試驗常無法正確預測藥物對人類的致畸胎性，且多數藥品上市前臨床試驗排除懷孕婦女。因此，為了取得更多的資訊以提供懷孕婦女更適當之用藥建議，藥品上市後研究 (如藥品處方型態研究及藥品風險研究) 至關重要。
研究目的：本研究探討臺灣懷孕婦女之藥品處方型態與相關不良事件。第一部分分析懷孕婦女整體用藥型態，以利指引後續藥品風險分析的研究方向。第二部分選定安胎藥品ritodrine為研究主題，分析其歷年使用趨勢以及藥物安全管理措施對使用型態之影響。第三部分則延續前一主題，探討使用ritodrine與懷孕婦女發生心肺不良事件之相關性。
研究方法：本研究之資料來源為臺灣之全國出生通報檔及全人口健康保險資料檔。主要研究族群為於出生通報檔有生產紀錄之懷孕20週以上的懷孕婦女 (懷孕期間以懷孕週數和新生兒出生日期定義)，於不同部分的研究另納入健保資料檔中15至49歲之育齡婦女或懷孕小於20週之流產婦女作為對照組。第一部分針對懷孕婦女與流產婦女分別分析其用藥型態，研究藥品範圍涵蓋所有非外用之藥品，並採用WHO ATC classification進行藥品分類。統計分析之指標為懷孕婦女於懷孕前84天、第一孕期、第二孕期、第三孕期、以及生產後84天使用藥品之比例以及使用頻率最高之藥品，並分析前述指標隨年代變化之趨勢。第二部分針對安胎藥品ritodrine，分析其歷年使用趨勢、處方特性、以及處方合理性，並比較2014年我國食品藥物管理署採行藥品安全管理措施前後之處方型態是否改變。第三部分則探討懷孕婦女使用ritodrine發生肺水腫、心律不整與心衰竭等心肺不良事件之風險，包含絕對風險 (以育齡婦女為標準人口分析心肺不良事件之年齡標準化發生率) 以及相對風險 (採用巢式病例對照研究，病例組為懷孕期間發生心肺不良事件者，配對相同懷孕天數時未發生事件且符合配對條件之對照組，以條件邏輯式迴歸模型分析勝算比 [aOR] 與95%信賴區間 [CI])。
研究結果：首先，我國懷孕婦女使用藥品之比例相當高 (2017年高達95.9%之孕婦曾於懷孕期間使用藥品)，且第二孕期與第三孕期之用藥比例有隨年代上升之趨勢。研究結果篩選出數個藥品值得進行後續研究，包含懷孕期間使用頻率最高之藥物 (progesterone與ritodrine)、流產婦女於第一孕期使用比例高於未流產婦女之藥物 (NSAIDs) 以及使用頻率高但過去懷孕風險分級為具有風險或風險資料不足之藥物 (pseudoephedrine與domperidone)。其次，本研究觀察到懷孕婦女使用ritodrine之比例隨著年代明顯上升，此趨勢主要來自口服劑型之使用率由2002年11.0% 增加至2018年達25.4%；反之，注射劑型之使用率於2010年以後微幅下降，至2018年為5.2%。然而，2014年實施藥品安全管理措施前後之ritodrine處方型態 (包含藥品使用天數、禁忌症族群使用情況等) 並未明顯改變。最後，本研究結果顯示，懷孕婦女於懷孕期間之心肺不良事件發生率並不高，但使用ritodrine之懷孕婦女相較於未使用者其心肺不良事件之風險顯著增加 (aOR 2.18 [95% CI 1.72–2.77])。其中，使用ritodrine注射劑型者心肺不良事件之風險較高 (肺水腫：aOR 10.56 [95% CI 6.39–17.45]；心律不整：aOR 4.15 [95% CI 1.99–8.64]；心衰竭：aOR 5.58 [95% CI 2.27–13.74])；使用ritodrine口服劑型者則仍有風險但風險較低 (肺水腫：aOR 1.76 [95% CI 1.12–2.76]；心律不整：aOR 2.21 [95% CI 1.47–3.32]；心衰竭：aOR 0.89 [95% CI 0.50–1.58])。
結論：本研究之特色在於針對我國藥品安全相關實證資料相當缺乏之懷孕婦女族群，採用具全國代表性且可提供完整懷孕相關資訊之資料檔，分析其用藥型態與相關藥品不良事件。研究結果有助於指引後續臺灣懷孕婦女藥品相關研究之研究方向，並可作為我國衛生主管機關執行藥品風險溝通與管理之參考。

關鍵字

懷孕婦女 ； 藥品處方型態 ； 藥品不良事件 ； 安胎藥品

並列摘要

Background: Evaluations of the benefits and risks of drugs used by pregnant women are very challenging. These drugs may have maternal side effects and may transfer from mother to fetus through the placenta, leading to adverse outcomes, such as spontaneous abortions, stillbirths, or congenital defects. In addition, for women with chronic illnesses requiring long-term treatment, discontinuation of drug therapy during pregnancy may impose risks to both the mother and fetus. Previous studies have shown that a large proportion of women take drugs during pregnancy and that the prevalence of drug use during pregnancy has increased over time. However, information on drug-related maternal or fetal adverse outcomes is scarce because most of the premarketing trials excluded pregnant women. Therefore, postmarketing studies using real-world data are crucial to support clinical practice and policy making regarding pregnant women.
Objectives: The overall aim of this study was to assess the use of drugs and associated adverse outcomes among pregnant women in Taiwan, and the study is divided into the following three subsections: (i) an examination of the utilization patterns of all drugs among pregnant women; (ii) an evaluation of prescription patterns of the tocolytic drug ritodrine with an exploration of whether the safety regulations issued by the Taiwan Food and Drug Administration (FDA) in 2014 affected its use; and (iii) a quantitation of the absolute and relative risks of maternal cardiopulmonary events associated with ritodrine.
Methods: This is a retrospective, nationwide study using data from two national databases, namely, the National Birth Certificate Application (BCA) database and the National Health Insurance Research Database (NHIRD). The primary study population was pregnant women above 20 weeks of gestation with records in the BCA database, and the duration of pregnancy was estimated using gestational age and the birth date of the newborn. Women of childbearing age and those whose pregnancies resulted in abortions were included as the comparison groups. In the first part of the study, the utilization patterns of all drugs among pregnant women (including the prevalence of drug use and the most frequently prescribed drugs) were assessed, and the utilization patterns between different pregnancy periods (before pregnancy, each trimester during pregnancy, and after delivery), different calendar years, and pregnant women with different pregnancy outcomes were compared. In the second part of the study, the annual prevalence of ritodrine use, as well as the characteristics and rationality of the ritodrine prescriptions, were assessed, and a before–after design was adopted to compare the utilization patterns of ritodrine before and after the safety regulation. In the third part of the study, the age-standardized incidence rates of maternal cardiopulmonary events (pulmonary edema, arrhythmia, and heart failure) were obtained using age distribution of childbearing-age women in Taiwan as the standard. Then, a nested case–control study design was adopted to examine the relative risk. Cases were defined by inpatient diagnoses of cardiopulmonary events plus prescriptions of specific medications, and controls were identified using risk set sampling. Adjusted odds ratios (aORs) and 95% confidence intervals (CIs) were estimated using conditional logistic regression models.
Results: In the first part of the study, the use of drugs during pregnancy was found to be common in Taiwan (95.9% of the pregnant women received drug prescriptions during pregnancy in 2017), and the prevalence of drug use in the 2nd and 3rd trimesters showed increasing trends from 2005 to 2017. Several drugs warrant future investigation, including the drugs most frequently prescribed during pregnancy (ritodrine and progesterone), drugs with a higher prevalence among pregnancies resulting in abortions than among those resulting in births (NSAIDs), and drugs commonly prescribed but classified as “human data suggest risk” or “limited human data” (pseudoephedrine and domperidone). In the second part of the study, the prevalence of using oral ritodrine was found to increase 2-fold from 2002 to 2018 (11.0% to 25.4%), while the prevalence of using ritodrine injection slightly declined after 2010 (5.2% in 2018). Although the Taiwan FDA issued a safety regulation in 2014, the regulatory action did not translate into more rational use of ritodrine among pregnant women (i.e., the prevalence of use, the treatment duration, and the prescription characteristics did not show a significant change after the regulatory action). In the third part of the study, although the incidence rates of cardiopulmonary events among pregnant women were low, pregnant women treated with ritodrine for tocolysis were at significantly increased risks of cardiopulmonary events compared to those without such treatment (aOR 2.18 [95% CI 1.72–2.77]). Specifically, exposure to oral ritodrine was associated with a lower risk of pulmonary edema (aOR 1.76 [95% CI 1.12–2.76]), arrhythmia (aOR 2.21 [95% CI 1.47–3.32]), and heart failure (aOR 0.89 [95% CI 0.50–1.58]), whereas exposure to ritodrine injection was associated with a significantly higher risk (corresponding aORs [95% CI]: 10.56 [6.39–17.45], 4.15 [1.99–8.64], 5.58 [2.27–13.74]).
Conclusions: This study utilized nationwide representative databases with detailed information on pregnancies to assess the utilization and adverse events of drugs among pregnant women, which is a highly understudied population in Taiwan. Findings from this study can provide insights for future studies on pregnant women and can support drug safety communications and regulations for the use of tocolytic drugs.

並列關鍵字

pregnant women ； drug prescribing patterns ； adverse drug events ； tocolytic drugs

參考文獻

1.Egen-Lappe V, Hasford J. Drug prescription in pregnancy: analysis of a large statutory sickness fund population. Eur J Clin Pharmacol. 2004;60(9):659-666.

2.Palmsten K, Hernández-Díaz S, Chambers CD, et al. The Most Commonly Dispensed Prescription Medications Among Pregnant Women Enrolled in the U.S. Medicaid Program. Obstet Gynecol. 2015;126(3):465-473.

3.Larcin L, Lona M, Karakaya G, Van Espen A, Damase-Michel C, Kirakoya-Samadoulougou F. Using administrative healthcare database records to study trends in prescribed medication dispensed during pregnancy in Belgium from 2003 to 2017. Pharmacoepidemiol Drug Saf. 2021;30(9):1224-1232.

4.Engeland A, Bjørge T, Klungsøyr K, Hjellvik V, Skurtveit S, Furu K. Trends in prescription drug use during pregnancy and postpartum in Norway, 2005 to 2015. Pharmacoepidemiol Drug Saf. 2018;27(9):995-1004.

5.Leong C, Chateau D, Dahl M, et al. Prescription medication use during pregnancies that resulted in births and abortions (2001-2013): A retrospective population-based study in a Canadian population. PLoS One. 2019;14(3):e0211319.