Clinical Characteristics Of Acinetobacter Baumannii And The Construction Of An Early Warning Model For Hospital Infection Risk

ObjectiveBy analyzing the characteristics of clinical distribution and changes in drug resistance of Acinetobacter baumannii（Ab）,we explore the independent risk factors for the occurrence of Ab,and establish an early warning model of infection risk based on these risk factors to provide a basis for the treatment and prevention and control of Ab infection.Methods1.Applying a case-control study,patients with Ab infection detected in the First Hospital of Jinzhou Medical University from January 2016 to December2019 were divided into a study group:634 patients with MDR infection and a control group:190 patients with non-MDR infection.WPS and SPSS software were used to statistically analyze the general condition,specimen composition,and departmental distribution of patients,and the change of drug resistance was analyzed by 2-test,and P<0.05 indicated that the difference was statistically significant.2.A self-made case information collection form for infected patients was used to collect relevant factors,including gender,age,smoking,alcohol consumption,APACHEII score,admission to ICU,cardiac insufficiency infectious shock,hypoproteinemia,surgery,artificial airway,central venous line placement,catheter placement,gastric tube placement,antibiotic use within one week before infection,hormone use within one week before infection,WBC,Hb,PLT,and other data.All relevant factors were subjected to t/2 test,and those with P<0.05 were included in the multifactorial logistic regression analysis to screen out independent risk factors for infection（P<0.05）,and a risk warning model was established,and its predictive efficacy was evaluated by Hosmer & Lemesho validation of the model goodness of fit,subject’s work characteristics ROC curve and area under the curve（AUC）.Results1.A total of Ab 824 strains were isolated in our hospital from 2016-2019,in which male patients were 1.83 of female patients.The age difference was large,concentrated between 50-89 years old,and the majority were 60-69 years old,and the average age of patients was 60.99 years.2.The departmental distribution of Ab was mainly in the intensive care unit（ICU）,neurosurgery,and respiratory medicine,with 40.0%,20.4%,and 13.2%,respectively.The highest detection rate among specimens was 84.0% for sputum specimens,5.7% for urine specimens,3.8% for blood specimens,and3.4% for secretion specimens.3.Ab whose 4-year resistance rates were 69.95% in 2016,70.65% in 2017,71.84% in 2018,and 76.93% in 2019,and the percentage of multi-drug resistant strains were above 74%.The resistance rate of methicillin/sulfanilamide was（45.0%-56.3%）,and the resistance rate of all other drugs was higher than 60%.Among them,the resistance rates of cefazolin,cefuroxime and cefoxitin were all higher than 96.0%,and the resistance rates of amikacin,tobramycin,cefotaxime and ceftazidime showed a significant increasing trend（P<0.05）.The mean resistance rate of meropenem was 71.1%,with a decreasing trend from 2016 to2018 and an increase in 2019.4.Independent risk factors for hospital-acquired infections in our MDRAB were APACHEII score（P=0.001）,WBC（P=0.009）,PLT（P=0.035）,hypoproteinemia（P=0.000）,artificial airway（P=0.000）,central venous line placement（P=0.002）,catheter placement（P=0.000）,gastric tube placement（P=0.001）,ICU admission（P=0.002）,cardiac insufficiency（P=0.000）,and antibiotic use within one week before infection（P=0.000）.5.Risk warning model:Logistic（P）=-7.954+2.154X₁+1.7X₂+1.597X₃+1.341X₄+1.157X₅+1.021X₆+1.080X₇+0.895X₈+0.049X₉+0.04X₁0-0.041X₁1.The use of antibiotics within one week before infection was recorded as X₁,artificial airway as X₂,catheter placement as X₃,ICU admission as X₄,hypoproteine mia as X₅,cardiac insufficiency as X₆,gastric tube placement as X₇,central venous line as X₈,WBC as X₉,APACHEII score as X₁0,and PLT as X₁1.The results of Hosmer & Lemeshow goodness of fit test were:chi-square=7.455,DF=8,P=0.488.Lemeshow goodness-of-fit test resultswere:chi-square=7.455,DF=8,P=0.488,the model fit was good（P>0.05）.The AUC under the R OC curve was 0.883,95% CI was 0.857-0.908,the sensitivity of the model was 83.0%,the specificity was 77.4%,the accuracywas 85.8%,the best cutoff value was 0.716,and the Jorden index was 0.604,which can be consid ered that this model has a good discriminatory effect and is clinically signi ficant.Conclusions1.Ab resistance in our hospital is serious,and the rate of drug resistan ce is increasing year by year,and it is heavy in the elderly.The infection m ainly invades the respiratory system,and the departments are widelydistribu ted,the most serious of which is ICU,followed by neurosurgery and respirat ory medicine.2.The independent risk factors for MDRAB hospital infection in our ho spital are:APACHEII score,WBC,PLT,hypoproteinemia,artificial airway,central venous placement,catheter placement,gastric tube placement,admission to ICU,cardiac insufficiency,and antibiotics used within one week before infecti on.3.A risk warning model was successfully established based on indep endent risk factors,and it can better predict the risk of MDRAB hospital inf ection,providing a basis for prospective clinical intervention.