Delivery of PrEP to Key Populations in Kenya: Applying Coincidence Analysis to Determine Effective Implementation Strategies

项目摘要 自2012年以来，暴露前预防（PrEP），通常作为每日药丸或最近作为双月一次 注射，已被用作预防艾滋病毒的一种方法。尽管PrEP在减少 为了降低一个人感染艾滋病毒的风险，采取了预防方法。在这项研究中，我们将研究 实施PrEP的现有计划，并评估哪些实施策略和背景因素导致 最佳的PrEP启动率和持续率;这些发现将有助于改善PrEP的提供 以及肯尼亚正在进行的艾滋病预防工作。本研究有两个目标：目标1）确定实施 影响PrEP在临床和社区实施的策略和背景因素 2）确定实施战略和环境战略的组合， 这些因素使PrEP最有效地提供给艾滋病毒感染风险最大的人群。到 为了实现目标1，我们将通过三个阶段收集和总结数据。在第1阶段，我们将请求聚合 来自肯尼亚开PrEP处方的组织的数据，以衡量研究结果（即，的比例 开始PrEP并在3个月和6个月期间维持PrEP使用的合格个人），以及 确定实施战略的初步变化和可能影响的现场层面的背景因素 PrEP实施。在第二阶段，我们将与工作人员在一部分研究中心进行定量调查， 实施研究综合框架因素的影响以及是否存在 关于PrEP实施成果的实施战略。在第三阶段，我们将完成半结构化 访谈，以更详细和细致入微地了解实施战略和背景 第二阶段反应的因素。为了实现目标2，我们将进行重合分析（CNA），以确定 导致符合条件的个人比例高的实施战略和背景因素 开始和维持PrEP的人（即，优化实现）。CNA是一种因果推理方法， 确定对于感兴趣的结果而言最小必要或足够的因素的组合。在 在这个应用程序中，我们将检查哪些上下文因素和实施策略的组合导致 优化PrEP实施。将根据以下三个结果运行模型： 启动PrEP的客户比例;以及3个月和6个月后继续使用PrEP的客户比例， 每一个都将被二分（高对低）或二分（高、中、低）。额外 代表客户群体差异的因素（例如，男男性行为者、妇女、青年）将被纳入 分析，以确定是否独特的策略和因素集导致不同的优化 人口。

Project Summary Since 2012, pre-exposure prophylaxis (PrEP), typically delivered as a daily pill or more recently as a bimonthly injection, has been available as a method of HIV prevention. Despite PrEP being highly effective at reducing a person’s risk of acquiring HIV, uptake of the prevention method has been. In this study we will examine existing programs implementing PrEP and assess which implementation strategies and contextual factors lead to the best rates of PrEP initiation and persistence; these findings will contribute to improving PrEP delivery and the ongoing HIV prevention efforts in Kenya. The study has two aims: Aim 1) Identify the implementation strategies and contextual factors affecting PrEP implementation within clinical and community-based organizations in Kenya and Aim 2) Determine the combination(s) of implementation strategies and contextual factors that produce the most effective delivery of PrEP to populations at greatest risk of HIV acquisition. To achieve Aim 1 we will collect and summarize data through three phases. In Phase 1 we will request aggregate data from organizations that prescribe PrEP in Kenya to measure the study outcomes (i.e., the proportions of eligible individuals who start PrEP and who have maintained PrEP use over 3-month and 6-month periods) and identify preliminary variation in the implementation strategies and site-level contextual factors that may affect PrEP implementation. In Phase 2 we will conduct quantitative surveys with staff at a subset of sites to assess the impact of Consolidated Framework for Implementation Research (CFIR) factors and presence or absence of implementation strategies on PrEP implementation outcomes. In Phase 3 we will complete semi-structured interviews to gain a more detailed and nuanced understanding of implementation strategies and contextual factors from Phase 2 responses. To achieve Aim 2 we will conduct Coincidence Analysis (CNA) to determine the sets of implementation strategies and contextual factors leading to high proportions of eligible individuals who both start and maintain PrEP (i.e., optimized implementation). CNA is a method of causal inference and determines the combinations of factors that are minimally necessary or sufficient for an outcome of interest. In this application we will examine which combinations of contextual factors and implementation strategies result in optimized PrEP implementation. The models will be run for three outcomes based on the following: proportion of clients who initiate PrEP; and proportion of clients continuing PrEP use after 3 and 6 months, each of which will either be dichotomized (high vs. low) or trichotomized (high, medium, low). Additional factors, representing differences in client populations (e.g., MSM, women, youth) will be included in the analyses to determine whether unique sets of strategies and factors lead to optimization for different populations.