基于贝叶斯网络的供应-加工-配送医用耗材库存预测与动态补货策略优化研究OA

Objective:To construct a management mode about supply-processing-delivery(SPD)inventory of medical consumables based on Bayesian networks,and verify application value of that in management for inventory of medical consumables.Methods:Main factors of influencing medical consumables were screened to construct a node set of Bayesian networks.The initial scheme sets of replenishment strategy were generated by using Whale Optimization Algorithm(WOA),and parameters training and constraint correction.The iteration and optimization were implemented according to the iteratively updated scheme of the WOA mechanism,and then,the strategy about prediction and dynamic replenishment for inventory of medical consumables was formed.The data of inventory of a total of 1,252 batches of medical consumables that were purchased by Pudong Branch of Shanghai Ninth People's Hospital,Shanghai Jiaotong University School of Medicine during 2023 and 2024 were selected.The conventional management mode for inventory was adopted to manage 607 batches of medical consumables that were purchased during January and December 2023,and SPD management mode for inventory of medical consumables(SPD management mode)based on Bayesian networks was adopted to management 645 batches of medical consumables that were purchased during January to December 2024.The indicators of management quality for inventory of medical consumables of two kinds of modes were compared.A self-designed structure questionnaire was used to assess the satisfaction of medical staffs for the services of use and management of medical consumables.Results:The average turnover rate and accurate rate of inventory of medical consumables of adopting the SPD management mode were respectively(96.76±1.58)%and(98.67±0.43)%,all of which were significantly higher than those of adopting the conventional management mode for inventory,and the differences were statistically significant(t=7.413,9.315,P<0.05).The average out-of-stock rate of consumables,and attrition rate of expired consumables of the SPD management mode were respectively(2.37±0.51)%and(2.19±0.39)%,all of which were significantly lower than those of the conventional mode for inventory,and the differences were statistically significant(t=7.277,10.631,P<0.05).The satisfaction scores of the quality reliability,variety diversity,adequacy of quantity,timeliness of response,and collaborative convenience of consumables that were managed by using the SPD management mode were higher than those by using the conventional management mode for inventory,and the differences were statistically significant(t=2.532,3.272,2.345,3.141,3.737,P<0.05).Conclusion:The SPD management mode based on Bayesian networks for inventory of medical consumable can enhance the management effect for inventory of medical consumables,and improve turnover efficiency of inventory,and increase accuracy of management for inventory,and reduce risks of out-of-stock and attrition,and enhance satisfaction of medical staffs for service quality of inventory management for medical consumables.