GOALI/Collaborative Research: Improving Patient Flow in Hospitals

这项学术与行业联络资助机会（GOALI）计划奖将通过减少医院内和医院之间的拥堵和改善患者流动来促进国民健康。通过开发预测流量动态和解决拥堵来源的模型，该项目将提高患者的安全性和满意度，同时减少医院的人员配备和设备成本。减少拥堵也将改善入住率预测，从而减少医院网络内的救护车改道。这项研究采用了新颖的建模技术，这些技术是由pi和GOALI合作伙伴西北纪念医院（NMH）密切合作得到的。将开发易于校准和足够通用的模型，以便其分析和见解适用于医院内广泛的临床环境。该奖项为研究生从事对医疗保健业务有直接好处的研究提供支持。pi将把他们的研究成果纳入排队和服务工程课程。研究结果将传播给运筹学研究和医疗保健从业者社区。该奖项支持多服务器排队模型的研究，该模型专门为医院操作量身定制，解决了患者流动动力学的关键特征。这些特征包括由于周期性放电窗口和非平稳到达过程而导致的突然离开。由于队列动力学的复杂性，本研究基于高峰负荷时的流体近似和轻负荷时的无限服务器近似，开发了适合队列现象的近似。这些方法与高保真度模拟相结合，将用于表征患者流动过程的长期行为，例如医院单元的最大吞吐量，以及它们在不同操作策略下的（周期性）时间依赖平衡行为。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This Grant Opportunity for Academic Liaison with Industry (GOALI) Program award will advance the national health by reducing congestion and improving patient flow within hospital units and between hospitals. By developing models that predict flow dynamics and address sources of congestion, this project will improve patient safety and satisfaction, and at the same time reduce staffing and equipment costs to the hospital. Reduced congestion will also improve occupancy prediction, thus reducing ambulance diversions within the hospital network. This research employs novel modeling techniques that are informed by a close collaboration between the PIs and the GOALI partner, Northwestern Memorial Hospital (NMH). Models will be developed that are easy to calibrate and sufficiently general, so that their analyses and insights are applicable in a broad range of clinical settings within the hospital. The award provides support for graduate students to engage in research that will be of direct benefit to healthcare operations. The PIs will incorporate the results of their research in courses on queueing and service engineering. The outcomes of the research will be disseminated to both the operations research and the healthcare practitioner communities.This award supports research on multi-server queueing models, tailored specifically to hospital operations, that address key features of patient-flow dynamics. These features include bursty departures due to periodic discharge windows and non-stationary arrival processes. Because of the complexity of the dynamics, the research develops appropriate approximations for queueing phenomena based on fluid approximations during peak loads and infinite server approximations during lightly loaded periods. These methods, in conjunction with high-fidelity simulations, will be used to characterize long-run behavior of the patient-flow process, such as the maximum throughput of hospital units, and their (periodic) time-dependent equilibria behavior under different operational policies.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.