| Objective To investigate the clinical characteristics of infected rheumatoid arthritis(RA)patients,and develop a rule to predict the infection.Methods The clinical data of 513 RA patients admitted to the Third Affiliated Hospital of Soochow University from December 2016 to December 2018 were collected.All the cases were divided into the non-infection group(n=330 patients)and the infection group according to whether they were infected or not,the severe infection group have 40 cases.The clinical data were analyzed by usingChi-square test,t-test,and Mann-Whitney U test.Using logistic regression we estimated the independent risk factors of infection in RA patients.Partial regression coefficient β of each risk factor was used to build the predictive rule.The performance of the prediction rule was assessed by AUC。Results(1)In total,513 RA patients were studied,the non-infection group has 330 patients,the infection group has 183 patients,the serious infection group has 40 patients,5 patients were dead.The most common infection sites are the respiratory system(61.20%),the urinary system(23.50%),skin and soft tissue(3.83%);67 patients with positive pathogenic bacteria were detected,with a positive rate of 36.61%.A total of 84 pathogenic bacteria were detected,of which 43 were bacteria,7 were mycoplasma,19 were fungi,and 11 were mixed bacteria;(2)Infection group patients differed significantly from the non-infection group with age,disease duration,disease activity(DAS28),white blood cell count,hemoglobin,platelet count,lymphocyte percentage,plasma albumin,erythrocyte sedimentation,C-reactive protein,CCP positive.Furthermore,hospitalization due to infection in the past year,long-term bed rest,diabetes,heart disease,chronic lung disease,chronic kidney disease,chronic liver disease,rheumatism,Sjogren’s syndrome,glucocorticoid treatment increased the risk of infection,whereas DMARDs reduced the risk.Serious infection group patients differed significantly from the non-infection group with:age≥70,diseaseduration≥10years,disease activity(DAS28),white blood cell count,hemoglobin,platelet count,percentage of lymphocytes,plasma albumin,erythrocyte sedimentation,C-reactive protein And anti-IgM antibodies.Furthermore,hospitalization due to infection in the past year,long-term bed rest,diabetes,heart disease,chronic lung disease,chronic kidney disease,rheumatism,Sjogren’s syndrome,glucocorticoid treatment increased the risk of infection,whereas methotrexate reduced the risk;(3)Multivariate logistic regression showed that disease activity(DAS28),hospitalization due to infection in the past year,diabetes and glucocorticoid treatment were independent risk factors for infection in RA patients(P<0.05);Age≥70years,disease duration≥10years,disease activity(DAS28),hospitalization for infection in the past year,heart disease,rheumatism and glucocorticoid treatment within the past year were independent risk factors for serious infection in RA patients(P<0.05),the results of the colinearity test showed that there was no multiple colinearity among the factors;(4)The dose of glucocorticoids is divided into four groups according to daily dose:no glucocorticoids,<7.5mg/day,7.6mg/day to 15.0mg/day,and>15.0mg/day.The risk of infection at different doses shows that as the dose of glucocorticoids increases,the risk of infection in RA patients increases.The dose of glucocorticoids is divided into one year’s cumulative dose:no Corticosteroids,>0.0 to 180.0mg,180.1mg to 1825.0mg,and>1825.0mg.The risk of infection in RA patients with different cumulative doses of glucocorticoids was compared.Increased,the risk of infection increased with the cumulative dose after exceeding 180.0mg;the DMARDs treatment was divided into three groups according to the current use:no DMARDs,1 DMARDs and ≥2 DMARDs,compared with different use The number of DMARDs-treated RA patients combined with the risk of infection,the results showed that when using one or more DMARDs,the risk of infection was not increased in RA patients;the treatment options for RA patients were divided into glucocorticoid without DMARDs,glucocorticoid only,glucocorticoid with 1 DMARDs,glucocorticoid with 2 DMARDs,the results show that both DMARDs and glucocorticoids alone Adding DMARDs on the basis,the risk of infection is lower than that of patients with glucocorticoid alone;(5)The predictive model of infection in RA patients is as follows:1 point for "duration of disease>10 years",4 points for "disease activity(DAS28)",3 points for "have been hospitalized for infection in the past two years",2 pointsfor "diabetes",2 points for "heart disease",1 point for "other rheumatism",1 point for "glucocorticoid treatment",-1 point for "methotrexate treatment".Total score>6 points can predict the occurrence of infection in RA patients in the future.The accuracy of the rule can be evaluated by area under the curve of receiver operating characteristic(ROC),the area under the curve is 0.762,the maximum cutoff point of the Jordan index is 0.425,sensitivity=0.639,specificity=0.785,P value<0.001;The predictive model of severe infection in RA patients is as follows:1 point for "age≥70 years",1 point for "duration of disease≥10 years",3 points for "DAS28 activity",and 2 points for "have been hospitalized for infection in the past two years" 2 points for "heart disease",2 points for "other rheumatism",1 point for "glucocorticoid treatment".Total score≥5 points can predict the occurrence of severe infection in RA patients in the future.The area under the ROC curve was 0.870,the maximum cutoff point of the Yoden index was 0.600,sensitivity=0.900,specificity=0.700,and P value<0.001.Conclusion The prediction rule showed good performance on identifying individuals at a higher risk of infection in RA patients,but it still needs further clinical validation in a lager population. |
PDF Full Text Request