Cohort Study of Opioids, Pain, and Safety in an era of Changing Policy (COPING)

 描述(由申请人提供)：我们提出了一项时间敏感的队列研究，研究对象是在改变阿片类药物处方政策的情况下接受阿片类药物治疗的慢性非癌性疼痛(CNCP)患者。美国在为CNCP积极开出阿片类药物几十年后，现在正在经历处方政策和做法的变化，预计这将减少许多患者获得阿片类药物的机会。旧金山健康网络(SFHN)是由旧金山公共卫生部(SFDPH)管理的综合诊所网络，其初级保健诊所和当地支付者也在2015年开始严格限制阿片类药物的处方。这些变化包括剂量限制，限制给有物质使用障碍(SODS)的人开处方，要求进行尿液测试，以及对处方药监测计划(PDMP)的检查。其中许多变化旨在减少分流，预计将导致许多患者终止阿片类药物治疗。这些变化对那些已经依赖阿片类药物的人的影响尚不清楚，尽管坊间证据表明，随着处方阿片类药物的终止，一些患者开始或恢复非法使用阿片类药物。此外，对于接受阿片类药物止痛的患者中阿片类药物过量的发生率、危险因素和事件水平特征还知之甚少。为了解决公共卫生研究中的这些重大差距，我们将建立政策变化(应对)时代阿片类药物、疼痛和安全性的队列研究，该研究将记录在处方阿片类药物供应变化的背景下疼痛、功能状态、处方阿片类药物使用和非法物质使用的变化，并确定处方阿片类药物处方人群中阿片类药物过量的模式和风险。我们将从SFHN慢性阿片治疗疼痛患者登记(疼痛管理登记，N=2,879)中招募400名处方≥每日相当于30毫克吗啡的阿片类药物治疗慢性非癌症疼痛的患者，包括至少100名积极使用非法阿片、可卡因或甲基苯丙胺的患者，以及至少100名艾滋病毒携带者。应对参与者将每6个月接受一次静脉采血；对阿片类药物、酒精和药物使用、疼痛控制和过量事件的行为评估(目标1)；艾滋病毒治疗结果(目标4)；以及功能状态的临床评估(目标2)。将进行两年一次的病历和行政数据审查，以评估重大医疗事件的发生率(目标3)。我们将使用广义估计方程(GEE)线性、Logistic和多项式模型来估计和比较行为和临床结果以及功能状态、处方阿片类药物(继续与减少/失去)以及总体趋势。我们将使用人-时间方法和具有稳定的治疗权重反向概率的边际结构模型来评估阿片类药物过量的发生率，以估计受先前暴露影响的时间依赖混杂因素病例以及受先前过量事件影响的暴露影响。

 DESCRIPTION (provided by applicant): We propose a time-sensitive cohort study of patients receiving opioid therapy for chronic non-cancer pain (CNCP) in the setting of shifting opioid prescribing policies. The United States, after decades of aggressive prescribing of opioids for CNCP, is now undergoing changes in prescribing policy and practice that is expected to reduce access to opioids for many patients. The primary care clinics of the San Francisco Health Network (SFHN), a network of integrated clinics managed by San Francisco's Department of Public Health (SFDPH), and local payers are also initiating tight restrictions on opioid prescribing in 2015. These changes include dose limits, restrictions on prescribing to persons with substance use disorders (SUDs), and required urine testing, and checks of the prescription drug monitoring program (PDMP). Many of these changes are designed to reduce diversion and are expected to lead to termination of opioid therapy for many patients. The impact of these changes on those already reliant upon opioids is unknown, although anecdotal evidence suggests that some patients initiate or resume illicit opioid use as prescribed opioids are terminated. In addition, there is little understanding of the rate, risk factors and event-level characteristics of opioid overdose among patients receiving opioids for pain. To address these major gaps in public health research, we will establish the Cohort study of Opioids, Pain and safety IN an era of chanGing policy (COPING), which will document changes in pain, functional status, prescribed opioid use and illicit substance use in the setting of shifting availability of prescribed opioids and identify opioid overdose patterns and risks in a population prescribed prescription opioids. We will recruit 400 patients prescribed ≥30 morphine equivalent milligrams of opioids daily for chronic, non-cancer pain, including at least 100 actively using illicit opioid, cocaine or methamphetamine and at least 100 living with HIV, from the SFHN registry of patients on chronic opioids for pain (Pain Management Registry, or "PMR", N=2,879), including patients whose opioids were recently discontinued. COPING participants will be seen every 6 months for phlebotomy; behavioral assessments on opioids, alcohol and drug use, pain control, and overdose events (Aim 1), and HIV treatment outcomes (Aim 4); and clinical evaluations for functional status (Aim 2). Biannual reviews of medical charts and administrative data will be conducted to evaluate incidence of major medical events (Aim 3). We will use generalized estimating equation (GEE) linear, logistic, and multinomial models to estimate and compare trends in behavioral and clinical outcomes and functional status, to prescribed opioids (continued versus reduced/lost access), and overall. We will assess incidence rates of opioid overdose using person-time methods and marginal structural models with stabilized inverse probability of treatment weights to estimate exposure effects for cases with time-dependent confounders that are affected by previous exposures as well as exposures that are affected by previous overdose events.