Readmission Risk Prediction Of Patients With Rheumatoid Arthritis Based On Medical Record Data

With the rapid development of machine learning and big data technologies,data mining has been widely used in the medical field.Clinical prediction based on data mining technology can effectively improve early warning and monitoring of clinical adverse events.Combined with data statistics and literature analysis,it is found that more and more patients with rheumatoid arthritis are admitted to hospital repeatedly,which has a great negative impact on society,hospitals,individual patients and their families.Therefore,this study was carried out to explore the risk prediction for readmission of patients with rheumatoid arthritis.A total of 899 inpatients with rheumatoid arthritis as the main cause were identified from the First Affiliated Hospital of Anhui University of Chinese Medicine between May 2012 and June 2015.The patient's demographic information,disease information,and laboratory indicators were extracted from the hospital's electronic medical record database.In this study,patient characteristics were modeled by using multivariate logistic regression method to explore the potential influencing factors for readmission and their degree of influence.Finally,the risk factors for readmission of patients with rheumatoid arthritis were confirmed by clinical theoretical support and data analysis.Of the patients with rheumatoid arthritis,15.68%had readmission records within one year,mean age was 52.9 years,mean length of stay was about 20 days and most patients were female(86.87%).The final multivariate logistic regression model included characteristics such as gender,age,length of stay,traditional Chinese medicine patterns,multiple laboratory indicators and comorbidities.Factors associated with increased risk of readmission included:female(odds ratio,OR=1.970,p=0.040),increased age(OR=1.649,p=0.023),increased length of stay(OR=1.883,p=0.018),rheumatoid factors positive(OR=1.638,p=0.049),anti-streptomycin O positive(OR=3.245,p=0.003),low hemoglobin levels(OR=1.523,p=0.035),comorbidities with respiratory disease(OR=2.868,p=0.047)and liver disorder(OR=7.041,p<0.001).Patients with Wind-damp-heat pattern had the highest risk of readmission,comparing to patients with other three traditional Chinese medicine patterns.To the best of our knowledge,this is the first study to explore risk factors for the readmission of patients with rheumatoid arthritis,filling the gaps in the relevant research fields.Through data mining of medical records,we found that anemia is an important risk factor affecting the readmission of patients with rheumatoid arthritis,which has a very important guiding significance for clinical research,disease diagnosis and treatment.