基于贝叶斯中断时间序列的抗菌药物管理措施效果评价OA

Objective:To explore the dynamic change characteristics of the incidence of hospital infections before and after the intervention of antimicrobial stewardship measures,and to quantitatively evaluate the short-term and long-term impacts of antimicrobial stewardship measures.Methods:Data on the infection rate of Klebsiella pneumoniae(KPN)among inpatients from July 2017 to December 2018 were collected as pre-intervention data.Data on the KPN infection rate among inpatients from January 2019 to June 2020 were collected as post-intervention data.The Bayesian interrupted time series method was used to evaluate the changing trend of KPN infection rate before and after the implementation of antimicrobial stewardship measures.Results:The average monthly KPN infection rate from July 2017 to June 2020 was 0.941%.The average monthly KPN infection rate before the implementation of antimicrobial stewardship measures was 0.947%.The average monthly KPN infection rate after the implementation of antimicrobial stewardship measures was 0.934%.The parameter estimation results of the model with a lag period of 2 months showed that(β)2 was statistically significant(P<0.05),with a risk ratio of 0.818.(β)3 was statistically significant(P<0.05),(β)1+(β)3=0.021,with a risk ratio of 1.021.(β)4 was statistically significant(P<0.05),with a risk ratio of 1.661.The results of the relative changes between the predictive values and counterfactual values showed that the changes from March to May 2019 all decreased(all P<0.05),while the changes from April to June 2020 all increased(all P<0.05).Conclusions:The Bayesian interrupted time series model could effectively evaluate the short-term and long-term effects of intervention measures.It could also evaluate whether the intervention had a lag effect or not.The implementation of hospital antimicrobial stewardship measures could reduce the infection risk in the short term,with a 2-month lag period in the intervention effect,while the long-term effect was not ideal.