Epidemiology And Control Study Of Gram-negative Bacteria In Intensive Care Unit Of A Tertiary General Hospital

Objective : Epidemiological characteristics and drug sensitivity of gram-negative bacteria(GNB)infections in the Intensive Care Unit(ICU)were analyzed to clarify the distribution and drug resistance of gram-negative bacteria in the ICU,and to construct a prediction model to predict the trend of Carbapenem-resistant gram-negative bacteria(CR-GNB)hospital infection in ICU.The model was developed to predict the trend of Carbapenem-resistant gram-negative bacteria(CR-GNB)hospital infections in intensive care units,and to provide a scientific basis for the formulation of reasonable prevention and control measures for Gram-negative bacteria infections in intensive care units and the rational use of clinical drugs.Method:Epidemiological characteristics and drug resistance of patients with GNB infection in 3826 clinical specimens from intensive care units in a tertiary general hospital in Inner Mongolia Autonomous Region from January 2020 to December 2022 were analyzed;22patient isolates of Acinetobacter baumannii in 2021 and 12 environmental isolates of Acinetobacter baumannii were collected for pulsed gel electrophoresis experiments and analyzed for homology;1105 cases of CR-GNB infection from 2020 to 2022 were selected Patients were analyzed for risk factors and logistic regression and decision tree prediction models were constructed,and the area under the ROC(AUC)was used to evaluate the discrimination of the models.Results:1.From 2020 to 2022,a total of 5626 strains were isolated in the intensive care unit of a tertiary general hospital in Inner Mongolia,including 3826 Gram-negative bacteria,with a detection rate of 68.01%,and GNB ranked first in the detection rate of pathogenic bacteria in ICU,with an overall increasing trend in the detection rate of GNB during the three years,among which the detection rate of CR-GNB among Gram-negative bacteria showed a decreasing trend,and CRAB and CRPA Among the Gram-negative bacteria,the detection rate of CR-GNB showed a decreasing trend,while the detection rate of CRAB and CRPA showed a decreasing trend;the comparison by year showed that a total of 1218 multi-drug resistant bacteria were detected in intensive care unit patients,and the detection rate of multi-drug resistant bacteria in CR-GNB intensive care unit showed an increasing trend year by year.2.1105 strains of CR-GNB were detected,and the main sources of specimens were sputum(71.86%)and lavage fluid(15.11%);the highest detection rate was in the lower respiratory tract(88.69%),followed by the urinary system(6.61%)and the hematological system(2.35%);the main distribution departments were neurosurgery(31.76%)and respiratory and critical care medicine(23.35%);the proportion of males was significantly higher than that of females.3.The resistance rate of CRAB to a variety of clinically used antimicrobial drugs such as cephalosporins is above 90%;the resistance rate of CRE to meropenem and imipenem is >90%;the resistance rate of CRPA to meropenem and furantoin is >95%..4.A total of 1268 ICU environmental specimens were collected,and the mean overall pass rate of the ICU environmental disinfection quality in this hospital was 94.47%,with statistically significant differences before and after intervention(ρ<0.05).The 34 strains of A.baumannii collected in2021 included 22 strains isolated from inpatients and 12 strains isolated from environmental specimens,and PFGE typing yielded a total of 7 clusters(A-J clusters)and 16 clonal types,and the similarity of all strains ranged from 52.43% to 100%,with the highest number of strains in cluster C.5.Risk factors for CR-GNB infection were explored by univariate and multifactorial analysis,and the results of multifactorial logistic regression analysis suggested that surgery during hospitalization(OR=5.051),malignancy(OR=2.448),ventilator use(OR=2.812)and antimicrobial drug combination(OR=3.865)were independent risk factor.Coma(OR=0.511)played a protective role against CR-GNB hospital-acquired infections.6,The AUC of the logistic regression model was 0.805(95% CI:0.772-0.839);the AUC of the decision tree model was 0.790(95% CI:0.756-0.824),and the AUC of both models was greater than 0.7,which was considered that both models The logistic regression model performed slightly better than the decision tree model.Conclusion: From January 2020 to December 2022,Gram-negative bacteria were the main pathogens in the intensive care unit of the hospital,and the composition ratio was increasing year by year.the male to female ratio of CR-GNB infection was 2.26:1,and it was predominantly in the elderly.Sputum specimens and lavage fluid are the main sources of CR-GNB specimens.The clinicians should choose antimicrobial drugs reasonably according to the results of drug sensitivity and patients’ clinical performance.The results of the homology analysis suggest that there is intra-hospital cross-infection of CR-GNB between patients and cleaning staff and between patients and patients in the hospital.Logistic regression models were slightly more effective than decision tree models in predicting CR-GNB hospital-acquired infections.