Study On The Influencing Factors Of In-hospital Infection In Patients With Multiple Myeloma And Construction Of Risk Assessment Model

Objective:This study retrospectively analyzed the clinical medical records of multiple myeloma patients hospitalized for chemotherapy in our hospital to explore the risk factors affecting the occurrence of infection in multiple myeloma patients,and to try to establish a prediction score model for multiple myeloma infection risk.Materials and methods:Clinical data of 100 hospitalized patients with multiple myeloma undergoing chemotherapy in the Department of Hematology,the First Affiliated Hospital of Jishou University from January 2017 to June 2020 were collected,to explore the influencing factors of infection after treatment.According to the ratio of 3:1,the above 100 multiple myeloma chemotherapy patients were randomly divided into the modeling group and the verification group according to the hospitalized cases(random seed: 20201214),the modeling data set is used to identify the risk factors,the Logistic regression model was constructed,according to the beta coefficient of each risk factor to carry on the assignment,create multiple myeloma infection risk prediction score model,using the receiver operating characteristic curve(ROC)evaluation model prediction accuracy,the data substituted into the validation group was evaluated again.Results:Of the 100 patients with multiple myeloma,625 times were hospitalized,211 times developed infection,with an infection rate of 33.76%,the lung was the main site of infection,and 160 cases developed pulmonary infection,with an infection rate of 25.60%.The influence factors of infection in patients with multiple myeloma analysis:Univariate factor analysis results suggest,lymphocyte count,c-reactive protein level,fasting blood glucose level,lactate dehydrogenase levels,percentage of bone marrow plasma cells,ECOG score,chemotherapy regimens is affecting multiple myeloma clinical important influence factors of hospital infection in patients with chemotherapy;Logistic regression multivariate analysis showed that C-reactive protein level,fasting blood glucose level,percentage of bone marrow plasma cells and ECOG score were independent risk factors for infection in hospitalized patients with multiple myeloma undergoing chemotherapy(P < 0.05).Analysis of the influence factors of lung infection in patients with multiple myeloma: Univariate factor analysis results suggest,lymphocyte count,c-reactive protein level,fasting blood glucose level,serum creatinine level,lactate dehydrogenase levels,ECOG score,international staging system,the DS in installment,the types of immune therapy,the existence of any type of catheter,chemotherapy regimens and its related;Logistic regression multivariate analysis showed that C-reactive protein level,fasting blood glucose level,serum creatinine level,ECOG score,DS stage and the presence of any type of catheter were independent risk factors for pulmonary infection in hospitalized patients with multiple myeloma undergoing chemotherapy(P < 0.05).Multiple myeloma patients in the modeling group were scored according to the established infection risk model,the risk score of infection in each multiple myeloma patient was calculated,and the infection rate corresponding to each score value was calculated.By observing the overall distribution of infection rate in multiple myeloma patients,patients were stratified according to the risk score: 0 ～ 2 were divided into low-risk group,with infection rate of 21.33%;3 ～ 5 were divided into moderate risk group,the infection rate was 43.45%;6 ～ 7 were divided into high-risk group,and the infection rate was 68.42%.The infection rate of each risk layer had statistical significance(P < 0.05),it is suggested that this model has a good function of risk prediction.Conclusion:1.C-reactive protein level,fasting blood glucose level,percentage of bone marrow plasma cells and ECOG score were independent risk factors for infection in hospitalized patients with multiple myeloma undergoing chemotherapy.2.C-reactive protein level,fasting blood glucose level,serum creatinine level,ECOG score,DS stage,and the presence of any type of catheter were independent risk factors for pulmonary infection in hospitalized patients with multiple myeloma undergoing chemotherapy.3.The infection risk prediction model established in this study has a good fitting degree and discrimination degree for the infection risk of multiple myeloma patients receiving in-hospital chemotherapy,and can be used to identify high-risk groups for early prevention and control.