Analysis Of Factors Of Antibiotic Use In Patients With AECOPD And Establishment Of Clinical Prediction Model

Objective:To retrospectively analyze AECOPD admitted patients,divide them into antibiotic use group and non-antibiotic use group,compare the clinical characteristics of the two groups,multivariate regression analysis of factors affecting the use of antibiotics in AECOPD patients,and model to predict whether AECOPD admission patients will initiate antimicrobial therapy,In order to reduce the exposure rate of antibiotics.Methods:Collect AECOPD patients who were hospitalized in the Second AffiliatedHospital of Kunming Medical University from January 1,2018 to December 31,2019.The ICD code is:J44.100.305 cases were enrolled and divided into AECOPD antibiotic use group and AECOPD non-antibiotic use group,of which 184 cases were in the antibiotic use group and 121 cases were in the non-antibiotic use group.Collect all patients’ gender,height,weight,body mass index(BMI,the calculation formula of BMI is:weight/height square(kg/m^2)),hospitalization time,hospitalization expenses,past medical history(including history of hypertension,diabetes History,history of respiratory failure,history of coronary heart disease,history of heart failure,history of asthma,etc.),smoking status(including history of smoking,whether you are still smoking,time of smoking),symptoms(whether there is purulent sputum,dyspnea,whether within 24 hours of admission Fever),whether to use antibiotics,whether to use antibiotics in combination,time of antibiotic use,history of mechanical ventilation,sputum culture,blood gas(PH,PCO2,PO2,CHCO3,BE)on the day of admission,lung function(FEV1,FVC,FEV1%pred,GOLD Grading percentage),blood cell analysis(WBC,EO,N,EO%,N%),PCT,CRP.Compare the differences between the two groups of patients.The AECOPD patients who used antibiotics were divided into GOLD1-2 group and GOLD3-4 group according to the GOLD classification,and the differences between the two groups were analyzed.Take all the collected variables as independent variables,and whether to use antibiotics as dependent variables are included in the logistic regression univariate analysis,select all independent variables step by step forward and backward regression for multivariate analysis to screen independent variables,and then conduct statistically significant independent variables Significance test.Finally,variables with significant results are included in the final logistic regression results.A nomogram is drawn to visualize the logistic regression results.The discrimination is used to check the ability to distinguish between the two groups of patients.Bootstrap re-sampling 1000 times internally validates the logistic regression results to evaluate the good fit inferior.The logistic regression results were used to establish model 1 for the independent variables,dyspnea,purulent sputum,and mechanical ventilation were used to establish model 2 for the independent variables,and the decision curve was used to compare the pros and cons of the two models.Result:1.Compared with the non-antibiotics group in AECOPD patients,the antibiotic group is older than the non-antibiotic group,and the difference is statistically significant(p<0.05);the proportion of dyspnea,fever,purulent sputum,and pulmonary heart disease in the antibiotic group It is higher than that of the non-antibiotic group,and the difference is statistically significant(p<0.05);the positive rate of sputum culture,gender,smoking,still smoking,diabetes,respiratory failure,heart failure,asthma,coronary heart disease,smoking time in the two groups There was no statistically significant difference(P>0.05).2.The difference of FEV1%pred between the antibiotic group and the non-antibiotic group was statistically significant(p<0.001),and the FEV1%pred non-antibiotic group was higher than the antibiotic group.3.The length of hospital stay was statistically significant between the antibiotic group and the non-antibiotic group(p<0.001).The antibiotic group was 8.00[7.00,10.00]compared with the non-antibiotic group 7.00[6.00,8.00].The antibiotic group 9600.85[8155.29,11822.25]was higher than the non-antibiotic group 6974.33[5838.52,8277.86].4.The percentage of white blood cells,neutrophils,neutrophils,PCT,CRP,PH,PCO2,CHCO3,BE in the antibiotic group were higher than those in the non-antibiotic group,the difference was statistically significant(p<0.05);There was no statistically significant difference in PO2 between the two groups(p>0.05).5.Among the patients taking antibiotics,the patients in the GOLD3-4 group were younger than those in the GOLD1-2 group,and the difference was statistically significant(p<0.05);the patients in the GOLD3-4 group had lower BMI than the GOLD 1-2 group The percentage of acidic granulocytes was lower,and the difference was statistically significant(p<0.05);compared with the GOLD 1-2 group,the percentage of respiratory failure and coronary heart disease in the GOLD3-4 group was higher,the hospitalization cost was higher,the number of neutrophils,and the number of neutrophils The percentage of acidic granulocytes,the percentage of neutrophils,the partial pressure of carbon dioxide,and the concentration of bicarbonate were higher,and the difference was statistically significant(p<0.05).6.P urulent sputum[adjusted OR:5.58,95%confidence interval:(2.28～15.4)],history of pulmonary heart disease[adjusted OR:2.54,95%confidence interval(1.39～4.59)],WBC[adjusted OR:1.24,95%confidence interval(1.07～1.45)],PCT[adjusted OR:31539,95%confidence interval(23.47～1.16e+08)],CRP[adjusted OR:1.04,95 confidence interval(1.00-1.09))],BE[adjusted OR:1.12,95%confidence interval(1.03～1.22)],dyspnea[adjusted OR:7.20,95%confidence interval(2.97～19.5)],mechanical ventilation[adjusted OR:0.34,95 confidence interval(0.18-0.62)]is a risk factor affecting the use of antibiotics in patients with AECOPD.A regression equation was constructed with risk factors of purulent sputum,pulmonary heart disease,WBC,PCT,CRP,BE,dyspnea,and mechanical ventilation,and named as model 1,logit(p)=-4.118+1.719 X purulent sputum+1.975 X dyspnea-1.082×mechanical ventilation+0.919×pulmonary heart disease+0.216×white blood cells+10.359×PCT+0.041×CRP+0.11×BE.The AUC of the area under the ROC curve of model 1 is 0.845,and the AUC of model 2 is 0.723.Model 1 is higher than Model 2 when the high-risk threshold(predicting the probability of using antibiotics)is 0.10-0.99,the net rate of return(using antibiotics can benefit).Conclusions:1.The antibiotic group had worse pulmonary function,older age,longer hospital stay,and higher hospitalization costs.The antibiotic group had a higher proportion of eosinophilic inflammation.2.Purulent sputum,pulmonary heart disease,WBC,PCT,CRP,BE,dyspnea,and mechanical ventilation are risk factors that affect the use of antibiotics in patients with AECOPD.3.Establishing a model with purulent sputum,pulmonary heart disease,WBC,PCT,CRP,BE,dyspnea,and mechanical ventilation as predictors can better predict whether patients with AECOPD need to start antibacterial therapy,and is better than purulent sputum and dyspnea,A model with mechanical ventilation as a predictor.