Research on obstetric ward planning combining lean thinking and mixed-integer programming

Abstract

Background: In recent years, there are many studies on scheduling methods of patient flow, nurse scheduling, bed allocation, operating room scheduling and other problems, but there is no report on the research methods of how to plan ward allocation from a more macroscopic perspective. Objective: Refine and stratify the obstetric ward to provide more accurate medical service for pregnant women and improve the work efficiency of obstetricians and midwives. The problem of how to allocate the number of each type of ward is modeled as a mixed integer programming problem, which maximizes the patient flow of pregnant women in obstetric hospitals. Methods: The obstetric wards are divided into observation ward, cesarean section ward and natural delivery ward according to lean thinking. CPLEX is used to solve the mixed-integer programming problem of ward allocation. In R software, multivariate Generalized Linear Models (GLM) regression model is used to analyze the influence of each factor on patient flow. Results: The maximum patient flow of each case was obtained by CPLEX, which was 19-25% higher than that of patients without refinement, stratification and planning. GLM regression analysis was carried out on the abovementioned data, and the positive and negative correlation factors were obtained. Conclusion: According to lean thinking, obstetric wards are divided into three types of wards. Obstetricians and midwives work more efficiently and get more rest time. Pregnant women also enjoy more detailed medical services. By modeling the delivery ward allocation problem as a mixed-integer programming problem, we can improve the capacity of the service in obstetric hospitals from a macro perspective. Through GLM regression model analysis, it is conducive to improve the obstetric hospital capacity from the perspective of positive and negative correlation factors.