Defining outcome measures for medication adherence in clinical trials.

偏离方案规定的给药方案，表现为对试验药物的不同依从性，是普遍和有问题的。一项对95项临床研究的分析报告称，服用处方口服药物的患者数量随着时间的推移逐渐减少，在100天后，只有不到70%的患者完全遵守了方案规定的给药方案[https://doi.org/10.1146/annurev-pharmtox-011711-113247]]。这可能导致对药物疗效的不正确解释，混淆适当给药方案的选择，并可能误导将安全问题归因于试验用药。考虑到注册试验中对药物依从性的测量差异很大，而且报道很少，这尤其成问题[https://doi.org/10.1002/cpt.2709]]。在国际药物依从性协会（ESPACOMP）的主持下，我们制定了药物依从性（ABC）的共识分类，药物依从性报告指南（EMERGE），分析依从性数据（TEOS）的方法以及评估依从性研究中的偏倚风险（RoBIAS和RoBOAS工具）。本博士项目旨在进一步推进临床试验中药物依从性测量和报告的方法。作为一个解释变量，对试验药物的依从性通常以服用剂量的比例来衡量（或在此基础上的一些变化）；并且，作为一个结果变量，在一段确定的观察期内，达到某一任意阈值（通常为80%）剂量的患者比例。这两种方法都掩盖了患者依从性本质上的重要差异。具体来说，非依从性包括非起始（这是一种二元结果）；给药方案执行不力（患者服药，但未按规定给药）；和过早停药（当他们未能坚持治疗时）。学生将：(i)回顾有关药物依从性的测量和度量的文献；（ii）利用来自数千名试验参与者的数据，评估这些措施与依从性三个阶段相关的偏倚风险，这些参与者使用药物事件监测系统（Aardex Group的依从性知识中心数据库）以电子方式测量依从性；（iii）在不同试验设计和目标的背景下评估不同措施的适当性；（iv）制定一套用于药物试验的核心结果依从性措施。

Deviations from protocol-defined dosing regimens, in the form of variable adherence to trial medication, are prevalent and problematic. An analysis of 95 clinical studies, reported that the number of patients taking prescribed oral medication(s) decreased progressively over time, with less than 70% of patients being fully adherent to the protocol-specified dosing regimen after 100 days [https://doi.org/10.1146/annurev-pharmtox-011711-113247]. This can lead to incorrect interpretation of a medicine's efficacy, confound the selection of an appropriate dosing regimen and may mislead the attribution of safety concerns to trial medication. This is especially problematic given that the measurement of medication adherence in registration trials varies widely and is reported poorly [https://doi.org/10.1002/cpt.2709]. Under the auspices of the International Society for Medication Adherence (ESPACOMP), we have developed a consensus taxonomy for medication adherence (ABC), medication adherence reporting guidelines (EMERGE), methods for analysing adherence data (TEOS) and for assessing the risk of bias within adherence research (RoBIAS and RoBOAS tools). This PhD project will aim to further advance the methodology of medication adherence measurement and reporting in clinical trials.As an explanatory variable, adherence to trial medication is conventionally measured as the proportion of doses taken (or some variation on this); and, as an outcome variable, as the proportion of patients achieving some arbitrary threshold (usually 80%) of doses taken over a defined period of observation. Both measures conceal important differences in the nature of patients' adherence. Specifically, non-adherence includes non-initiation (which is a dichotomous outcome); poor implementation of the dosing regimen (patients who take the drug, but not according to the prescribed dosing regimen); and premature discontinuation (when they are fail to persist with treatment). The student will: (i) review the literature for measures and metrics of medication adherence; (ii) assess the risk of bias of these measures in relation to the three phases of adherence, utilising data from several thousand trial participants in whom adherence was measured electronically using the Medication Event Monitoring System (Aardex Group's Adherence Knowledge Centre database); (iii) assess different measures for their appropriateness in the context of different trial designs and objectives; and (iv) develop a core outcome set of adherence measures for use in drug trials.