DAT-Emulating target trials with big data to strengthen the evidence base for the clinical management of opioid use disorder

摘要 DAT18-06 2016年，在阿片类药物相关过量死亡大幅增加后，不列颠 加拿大哥伦比亚(BC)宣布进入公共卫生紧急状态。不列颠哥伦比亚省报告2018年有1510人死于非法药物过量，这一比率 每10万人中有31.0人，在美国各州中排名第10，与罗德岛(31.0)相当，仅次于西弗吉尼亚州(57.8)、俄亥俄州(46.3) 宾夕法尼亚州(44.3)，但领先于新泽西州(30.0)，密歇根州(27.8)和佛罗里达州(25.1)。而四倍的增长 自2012年以来观察到的过量死亡很大程度上可归因于芬太尼和其他合成阿片类药物的污染 在非法药物供应方面，许多死者过去曾寻求治疗，并复发。 阿片激动剂治疗(OAT)是抗击这一流行病的最有效工具之一。然而，燕麦片 在卑诗省和国际上，吸收和留存是次优的，主要受到社会和结构性因素的影响，但也 偏离循证护理标准和支离破碎的护理，特别是对那些并发疾病的人。 此外，加拿大和美国管理阿片类药物使用障碍的临床指南的一些方面(OUD) 都是基于有限和低质量的证据。即使是现有治疗相对有效性的基本证据 总体上和芬太尼时代的关键人群缺乏选择。 这项建议旨在应用尖端的因果推断方法来模拟三个阶段的一系列“目标试验”。 近期目标：(1)确定美沙酮与丁丙诺啡/纳洛酮治疗不同疾病的疗效比较 (2)确定尿液药物筛查的影响--这是一种普遍存在的非证据-- 基于OUD关怀的要素-关于燕麦滞留和死亡率；以及(3)确定燕麦对以下方面的补充作用 合并肝炎的阿片使用障碍(PWOUD)患者对定向作用抗病毒药物的吸收和依从性 C病毒。“靶向试验”框架传递了反事实理论，为比较效果提供了灵活的基础。 以意向治疗或按方案为基础的治疗和临床管理策略。我们已经确定了一个 与PWOUD的临床管理相关的额外问题的数量，其中指南得到了较弱的支持 经验证据，并将努力回答尽可能多的这些问题，使用共同的 框架侧重于方法的严密性和

ABSTRACT DAT18-06 In 2016, following a dramatic increase in opioid-related overdose deaths, the province of British Columbia (BC), Canada declared a public health emergency. BC reported 1,510 illicit drug overdose deaths in 2018, a rate of 31.0 per 100,000, placing it 10th among US states, equal to Rhode Island (31.0), behind West Virginia (57.8), Ohio (46.3) and Pennsylvania (44.3) but ahead of New Jersey (30.0), Michigan (27.8) and Florida (25.1). While the four-fold increase in overdose deaths observed since 2012 is largely attributable to the contamination of fentanyl and other synthetic opioids in the illicit drug supply, many decedents had sought treatment in the past and experienced relapses. Opioid agonist treatment (OAT) is among the most effective tools available to combat the epidemic. However, OAT uptake and retention is sub-optimal in BC and internationally, compromised largely by social and structural factors but also deviations from evidence-based standards of care and fragmented care, particularly for those with concurrent disorders. Furthermore, a number of aspects of Canadian and US clinical guidelines for the management of opioid use disorder (OUD) are based on limited and low-quality evidence. Even basic evidence on the comparative effectiveness of available treatment options overall and for key populations in the fentanyl era is lacking. This proposal aims to apply cutting-edge methods for causal inference in emulating a series of ‘target trials’ in three immediate aims: (1) to determine the comparative effectiveness of methadone versus buprenorphine/naloxone for different patient subgroups presenting for OAT; (2) to determine the impact of urine drug screening – a ubiquitous and non-evidence- based element of OUD care – on OAT retention and mortality; and (3) to determine the complementary effects of OAT on uptake and adherence to directing-acting antivirals for people with opioid use disorder (PWOUD) with concurrent Hepatitis C Virus. The ‘target trial’ framework channels counterfactual theory in providing a flexible basis for comparing the effects of treatment and clinical management strategies on either an intent-to-treat or per-protocol basis. We have identified a number of additional questions related to the clinical management of PWOUD where guidelines are supported by weaker empirical evidence and will endeavor to answer as many of these questions as possible, prioritizing analysis using a common framework focused on methodological rigor and