Using Machine Learning and Blockchain Technology to Reduce Drug Diversion in Hospitals

据分析，因分流造成的剂量损失量从2017年的2100万增加到4700万 2018年达到100万，增长126%。解决毒品分流问题是一个多方面的问题，包括 从提供商培训到硬件和软件系统的实施，再到 管理对受管制物质的获取。然而，尽管最近在控制和监测方面有所改进， 为了获得受控物质，查明毒品转移和确保遵约的过程是 既复杂又耗时。我们的总体目标是进一步开发基于区块链的技术 跟踪和记录医院环境中受控物质的运输和管理 并检测到毒品转移。该技术的可行性和相关的基于机器学习的数据 在一期工程中建立了分析引擎。我们的具体目标是：1.进一步发展 基于云的软件平台，利用智能手机捕获临床上的药品交易。我们会 进一步开发基于云的软件平台及其相关的智能手机应用程序和基于网络的 在第一阶段项目中开发的仪表板。软件平台，它使用区块链来创建 一成不变的审计跟踪，将被进一步开发来捕获受控的“从摇篮到坟墓”的文档 门诊外科中心内的物质使用和位置。在部署之前，该平台将 通过使用我们在第一阶段项目中开发的In Silico模型进行模拟测试，该模型能够 通过使用多代理模拟框架生成真实的交易数据。2.部署软件 合作卫生系统的平台。我们将分两个阶段推出我们的软件平台， 在第一阶段(这一特定目标的重点)，我们计划对我们的智能手机应用程序进行试点测试 以及相关的管理员基于Web的仪表板，以最终取代纸质日志。这样做的目的是 部署的阶段是了解和解决部署新系统的挑战和潜在影响 关于工作流程及其采用。与采用、遵守协议和对临床的影响相关的数据 将衡量工作流程，并通过对软件平台进行微调来应对任何挑战。3. 进一步开发和部署分析引擎，以检测毒品转移。这样做的目的是 具体目标是进一步开发和微调使用数据(记录在区块链上)的分析引擎 以侦测毒品转移。这项工作涉及两项主要任务。真实数据(作为具体目标2的一部分收集)将 由第一阶段开发的框架离线分析，通过与我们的协作健康 系统，我们将调查生成的危险信号，并使用此类调查的结果来微调 我们模型的参数。接下来，我们将集成更新的分析引擎，并推出毒品分流 在推广的第二阶段，与伙伴卫生系统合作，向飞行员提供检测能力。

Based on an analysis, the volume of dosage lost due to diversion increased from 21 million in 2017 to 47 million in 2018, a 126% increase. Addressing the drug diversion problem is a multi-faceted problem involving many components ranging from provider training to implementation of hardware and software systems to manage access to controlled substances. However, despite recent improvements in controlling and monitoring access to controlled substances, the process of identifying drug diversion and ensuring compliance is complicated and time consuming. Our overall goal is to further develop a technology based on blockchains to track and document transportation and administration of controlled substances in a hospital environment and detect drug diversion. The feasibility of the technology and the associated machine learning-based data analysis engine was established in the Phase I project. Our specific aims are: 1. Further Development of a cloud-based software platform to leverage smartphones to capture drug transactions in clinic. We will further develop the cloud-based software platform and its associated smartphone app and web-based dashboard developed in the Phase I project. The software platform, which uses blockchains to create an immutable audit trail, will be further developed to capture the “cradle-to-grave” documentation of controlled substance use and location within an ambulatory surgical center. Prior to deployment, the platform will be tested by simulation testing using the in silico model developed in our Phase I project, which is capable of generating realistic transaction data by using a multi-agent simulation framework. 2. Deploying the software platform at the collaborating health system. We will perform a two-stage rollout of our software platform, where in the first stage (the focus of this Specific Aim), we plan to perform a pilot test of our smartphone app and the associated administrator web-based dashboard to ultimately replace the paper logs. The goal of this stage of the rollout is to understand and address challenges of deploying a new system and potential impacts on the workflow and its adoption. Data related to adoption, adherence to protocols, and impact on clinical workflow will be measured and any challenges will be addressed by fine-tuning of the software platform. 3. Further development and deployment of an analysis engine to detect drug diversion. The goal of this specific aim is to further develop and fine-tune the analysis engine that uses data (recorded on the blockchain) to detect drug diversion. This effort involves two main tasks. Real data (collected as part of Specific Aim 2) will be analyzed offline by the framework developed in Phase I. Through collaboration with our collaborating health system, we will investigate the generated red flags and use the results of such investigation to fine-tune the parameters of our model. Next, we will integrate the updated analysis engine and rollout the drug diversion detection capability to the pilot in collaboration with the partner health system in the second stage of the rollout.