面向疫情防控的医院资源序贯优化配置

Sequential Optimization of Hospital Resources Planning for Controlling Epidemic

有限医疗资源的快速合理配置是控制疫情的有效手段。针对疫情发展不确定情况下的医院资源配置问题,首先,建立了具有独创性的传染病模型,直接量化了医院资源对疫情发展的影响;其次,提出了序贯决策框架及时更新疫情发展数据和资源决策结果,并基于该框架构建风险厌恶型的随机优化模型对资源进行合理配置;最后,通过基于武汉市新冠肺炎疫情的案例研究,证明了传染病模型能较好地预测疫情发展,序贯优化决策的资源配置结果能减少资源的使用数量,并更好更快地控制疫情。与风险中性的随机优化模型相比,风险厌恶的随机优化模型得到的资源配置结果较稳定且能更好地控制疫情,更具辅助实际决策的价值。

The rapid and appropriate allocation of limited healthcare resources is an effective way to control the spread of epidemic. Aimed at the problem of hospital resource allocation under the uncertain spread of epidemic,firstly,a newly introduced spread model was established to directly quantify the impact of hospital resources on the spread of the epidemic. Secondly,a sequential decisionmaking framework was proposed to update the epidemic data and resource allocation results in time,and a risk-averse stochastic programming was constructed to allocate resources under the uncertainty of the epidemic spread. Finally,based on the real world case of the COVID-19 outbreak in Wuhan,the validity of the epidemic spread model was verified. The proposed sequential decision-making framework and risk averse stochastic programming could control epidemic better and faster,and reduce the number of resources used. In addition,the risk-averse design of the objective was able to take a more steady approach to resource planning helped avoid large fluctuations in resource allocations compared to a risk-neutral design.