Operations Research for Hospital Efficiency and Effectiveness

这项研究计划的总体目标是为医院管理提高运筹学和管理学的最新水平。在这方面，效率是指利用经营医院所需的各种资源，例如医生、护士、专职医疗专业人员、技术员、护理员、工作人员；效力是指医院系统在成本、护理质量、获得护理的机会和健康结果方面的整体表现。效率和效力的提高涉及多个利益攸关方、多个目标、各种限制，因此需要进行一些艰难的权衡。这构成了具有挑战性的问题的温床，其中许多问题目前正由医院管理人员根据经验和直觉加以解决。运筹学和工业工程构成了方法论领域，通过开发可靠的分析框架、决策支持工具和管理见解，我们可以从这些领域中受益于改善当前的医院管理实践。 在这项拨款期间，我们会研究有关医院效率和成效的三个具体问题：(I)医院出院程序与社区护理的整合；(Ii)医院住院人数的策略管理；及(Iii)住院病房的运作管理。虽然临床环境，包括医院的类型和患者的疾病状况，会对所产生的数学模型的细节产生影响，但我们的目的是努力开发一个通用的建模框架，使我们能够获得普遍适用的管理见解。鉴于我们计划研究的问题的时间依赖性，模拟、随机动态规划和马尔可夫决策过程构成了我们计划深入研究的建模工具包。我们将开发近似动态规划算法或采用分解方法和元启发式框架来解决正在出现的大规模数学公式。 我们计划调查的具体问题包括：精神病患者从急症护理机构过渡到不同级别监管的住房单位(在(I)项下)、非服务病房分配对医院整体表现的影响(在(Ii)项下)，以及为外科病房的运作开发基于实时模拟的优化方法(在(Iii)项下)。我们在这三个方面的合作伙伴分别是道格拉斯医院、蒙特利尔综合医院和犹太综合医院(JGH)。特别是，将利用目前安装在JGH两个外科病房的实时定位系统(RTLS)开展关于业务决策的第三个研究流，这要归功于向申请人提供的CFI赠款。

The overarching objective of this research program is to improve the state of the art in operations research and management science for hospital management. In this context, efficiency refers to the utilization of the variety of resources that are required for running a hospital e.g., physicians, nurses, allied health professionals, technicians, orderlies, staff; and effectiveness refers to the overall performance of the hospital system in terms of the cost, quality of care, access to care and health outcomes. The improvement of efficiency and effectiveness involve multiple stakeholders, multiple objectives, a variety of constraints and hence, a number of difficult trade offs. This constitutes a hotbed of challenging problems many of which are currently being tackled by hospital administrators based on experience and intuition. Operations research and industrial engineering constitute methodology domains from which we can benefit in improving the current practice of hospital management by developing solid analytical frameworks, decision support tools and managerial insights. During the period of this grant, we will investigate three specific issues concerning hospital efficiency and effectiveness: (i) integration of hospital discharge processes with community-based care, (ii) tactical management of the inpatient capacity in a hospital, and (iii) operational management of inpatient wards. Although the clinical context, including the type of hospital and the patient’s disease condition, would have an impact on the specifics of the resulting mathematical models, our intent is to strive for the development of a generic modeling framework that would enable us to derive generally applicable managerial insights. In light of the time-dependent nature of the problems we plan to study, simulation, stochastic dynamic programming and Markov decision processes constitute the modeling toolkit we plan to dwell on. We will develop approximate dynamic programming algorithms or adopt decomposition methods and meta-heuristic frameworks for solving the arising large-scale mathematical formulations. The specific problems we plan to investigate include the transitioning of mentally ill patients from acute care facility to housing units with varying levels of supervision (under (i)), the impact of off-service ward assignments on the hospital’s overall performance (under (ii)), and the development of real-time simulation based optimization methods for the operation of surgical wards (under (iii)). Our partners for these three streams of research are the Douglas Hospital, Montreal General Hospital and Jewish General Hospital (JGH), respectively. In particular, the third research stream on operational decisions will be carried out utilizing the real-time location system (RTLS) that is now installed at two surgical wards of JGH, thanks to a CFI grant to the applicant.