Risk Factors Analysis And Prediction Model Of Primary Sjogren’s Syndrome Complicated With Interstitial Lung Disease

Objective This study aims to explore the risk factors of primary sjogren’s syndrome complicated with interstitial lung disease,establish a clinical prediction model,and provide reference for the early detection of primary sjogren’s syndrome complicated with interstitial lung disease.Methods Clinical data of patients diagnosed with primary Sjogren’s syndrome in the Department of Rheumatology,Affiliated Hospital of Nantong University from January 2013 to July 2021 were collected,including:1.General information: gender,age at the time of diagnosis of pSS,course of disease(from the first symptom to the time of treatment),smoking history,family history(whether there is a family history of rheumatism in the paternal or maternal line),hypertension,diabetes,history of tumor;Clinical symptoms: dry mouth,dry eyes,dental caries,parotid gland enlargement,joint pain,dry cough,chest tightness,shortness of breath,Reynolds phenomenon;Laboratory examination:Routine blood(white blood cell,platelet,hemoglobin,neutrophil count,lymphocyte count),routine urine(uric acid alkalinity,urine specific gravity),liver(nmda aminotransferase,alanine aminotransferase,lactate dehydrogenase and the ratio of albumin,white balls,kidney(creatinine),electrolyte,blood sedimentation,C-reactive protein(potassium),rheumatoid factor,immune Globulin(Ig G,Ig A,Ig M),complement C3,C4,ANA,anti-SSA antibody,anti-SSB antibody;Imaging data: LUNG HRCT;Pathological data: pathological results of labial gland biopsy.2.Patients were divided into two groups according to whether interstitial lung disease was associated with pSS-ILD and non-interstitial lung disease(pSS),and the differences in general characteristics,clinical manifestations,and laboratory tests were compared between the two groups.3.LASSO regression was used to screen risk factors that might be associated with pSS-ILD,and Logistic regression modeling was used to establish a clinical prediction model based on the line graph.The area under the ROC curve(AUC)was used to evaluate the distinguishing ability of the prediction model,and the accuracy of the prediction model was evaluated by drawing correction curves.Hosmer-lemeshow test was used to determine the goodness of fit of the model,and DCA curve was used to analyze its clinical practicability.Results A total of 389 patients with primary sjogren’s syndrome were included in this study,and 99 patients(25.45%)with pSS complicated with ILD were included.There were 29 males and 360 females,with a male to female ratio of 1:12.The age range of diagnosis of pSS ranged from 18 to 86 years,with an average age of(52±17)years at diagnosis of pSS.From the aspect of composition ratio,females accounted for a large proportion(92.54%);Most pSS patients without ILD(74.55%).1.Descriptive statistics and single factor analysis of clinical data of patients in the two groups showed that there was no significant difference in gender,duration of disease,family history,hypertension,diabetes and history of malignant tumor between the two groups.The age at diagnosis of pSS in the pSS-ILD group was(56±12)years older than that in the pSS group(50±16)years older.pSS-ILD patients are more likely to have dry cough,chest tightness,shortness of breath,dental caries clinical symptoms.In terms of laboratory examination,there were differences in ANA,anti-SSA /60 KD,white blood cell,platelet,neutrophil,albumin,white ball ratio,lactate dehydrogenase,blood potassium,ESR,C-reactive protein,complement C3 and urine p H between the two groups,and P values of all the above related variables were<0.05.2.A total of 99 patients with sjogren’s syndrome complicated with interstitial lung disease were included in this study.Among them,84 patients showed a single HRCT pattern,mainly grid-like changes,and 15 patients showed a mixed HRCT pattern,mainly grid-like changes and fiber strip shadow.3.Significant variables were screened by LASSO regression,and the variables with non-0 coefficient were eliminated,and 6 variables with0 coefficient were retained,including age at diagnosis of pSS,dry cough,shortness of breath,white blood cell,erythrocyte sedimentation rate and C-reactive protein,and then included in Logistic regression.It was found that age(P>0.05),erythrocyte sedimentation rate(P>0.05)and leukocyte(P>0.05)at the time of diagnosis of pSS did not pass the Logistic significance analysis test.According to the clinical analysis,leukocyte variable was removed from the model,and finally 5 risk factors related to pSS-ILD were screened as follows: age at diagnosis of pSS(OR: 2.780,95%CI: 0.765-9.359,P=0.106),dry cough(OR: 12.416,95%CI: 6.332-25.154,P<0.001),shortness of breath(OR: 6.294,95%CI: 2.757-14.889,P<0.001),ESR(OR: 1.505,95%CI: 0.724-3.306,P=0.288),C-reactive protein(OR: 4.870,95%CI: 2.035～11.712,P<0.001),were incorporated into the Logistic regression prediction model,and the prediction model was visualized by using the line chart.4.The evaluation of the line graph model showed that the AUC value of the prediction model for the occurrence of pSS-ILD was0.8414(95%CI:0.791,0.892),indicating that the model had a good discrimination ability,that is,hosmer-Lemeshow test showed that the prediction model had a good fit(P=0.812>0.05).DCA curve analysis showed that when the probability of ILD with pSS was in the range of10%-100%,the rosette had clinical validity.ConclusionsThe independent risk factors for primary sjogren’s syndrome complicated with interstitial lung disease were as follows: age greater than 70 years at diagnosis of pSS,dry cough,shortness of breath,increased erythrocyte sedimentation rate,and increased C-reactive protein.Based on the significant variables screened by LASSO regression,the linear graph prediction model constructed by Logistic regression analysis has good clinical application value in predicting whether patients with primary Sjogren’s syndrome are complicated with interstitial lung disease.