A Novel Predictive Model To Assess The Prognosis Of Patients With Severe Fever With Thrombocytopenia Syndrome And Treatment Evaluation

Background Severe fever with thrombocytopenia syndrome,first discovered in 2009,is a new infectious disease caused by the new Bunia virus,which is characterized by fever,fatigue,thrombocytopenia,leukopenia,gastrointestinal and central nervous system symptoms;The condition of SFTS changes rapidly with high mortality,so early prognosis is very important for the disease.Patients with SFTS lack characteristic clinical manifestations in the early stage,and the risk factors of death in patients with SFTS are not very clear.At present,several predictive models reported in the literature related to the prognosis of SFTS are all the single-center studies,lack of reliable verification data.This study intends to analyze the correlation between disease outcomes of patients and some affecting factors such as patients'clinical characteristics,laboratory parameters and virological indicators,and establish a predictive model related to disease severity and prognosis.Multicenter data were used to validate the stability of the model,and to evaluate the clinical instructive value of the model for treating the individual patient,providing a basis for early judgment of patient prognosis and optimization of treatment plans.Objective To establish a reliable model to evaluate the severity and prognosis of SFTS,and to explore the efficacy of various treatment schemes.Methods A retrospective analysis was made on 377 confirmed SFTS cases admitted to the second Hospital of Anhui Medical University,the first affiliated Hospital of Anhui Medical University and the first affiliated Hospital of China University of Science and Technology from June 2014 to December 2019.Among them,161 patients from the second Hospital of Anhui Medical University were selected as the model group,and a total of 216 patients from the first affiliated Hospital of Anhui Medical University and the first affiliated Hospital of China University of Science and Technology were selected as the verification group.Their clinical symptoms and signs,such as nervous system symptoms,digestive tract symptoms,respiratory symptoms,bleeding tendencies and so on,were observed.SFTSV nucleic acid was determined by fluorescence quantitative polymerase chain reaction and Enzyme-linked immunosorbent assay was used to detect STTS virus-specific Ig M and Ig G antibodies as well as blood routine,biochemistry blood coagulation and other general items;According to the disease outcomes,the patients were divided into survival group and death group.The differences of various indicators,clinical characteristics and pathogenic results between the two groups were analyzed.The independent risk factors affecting the prognosis of the disease were established according to Logistic regression.Based on these risk factors,the prognosis prediction model was established.According to the model,the patients were divided into high-risk,medium-risk and low-risk subgroups,and the disease outcomes of the three groups were compared.Meanwhile,the clinical characteristics and laboratory data of the patients in the verification group were analyzed.According to the established predictive model,the verification group was divided into high-risk,medium-risk and low-risk subgroups,and the disease outcomes of were compared in the three subgroups.The disease outcomes of the corresponding verification group and model group for each subgroup were compared respectively to verify the stability of the model.In addition,according to the therapeutic responses of patients in different subgroups to ribavirin,gamma globulin and antibiotics,the study analyzed the effects of these drugs on the prognosis of patients with different risk stratification,and further evaluated the guiding significance of the model to the treatment of patients.Results?1?There were no statistically significant differences in the basic clinical characteristics between the model group and the verification group,such as sex,age,WBC,N,L,RBC,PLT,Ca²⁺,TBIL,ALT,AST,CK,PT-INR,AMY,SFTSV RNA,etc,suggesting that the data sources of the model group and the verification group were balanced.?2?The results of univariate analysis in the modeling group showed that there were statistically significant differences in age,RBC,Hb,PLT,Ca,Glu,e GFR,ALT,AST,LDH,PT-INR,DD dimer,AMY,HPL,SFTSV specific Ig M antibody,SFTSV RNA,complicated with pancreatitis,gastrointestinal bleeding,cutaneous mucosal bleeding,CNS symptoms or urinary protein between the two groups.?The OR values were 1.054,0.425,0.973,0.963,0.023,1.122,0.979,1.003,1.001,1.001,74.745,1.094,1.001,1.000,0.343,2.028,4.655,8.205,6.042,168.000,5.261?;There was a strong correlation between CNS symptoms and death outcome.The results of Logistic regression showed that age,SFTSV RNA and gastrointestinal bleeding were independent risk factors affecting the prognosis of SFTS patients.According to the results of multivariate analysis,the formula was established:Score of joint index=age+10*SFTSV RNA?lg?+30*gastrointestinal bleeding,the best cut-off value was 118.The values of AUC,sensitivity and specificity of the score of the joint index were 0.859,0.962 and 0.677,respectively.?3?According to CNS symptoms and the score of the joint index,the patients were divided into three subgroups.The results showed that the mortality were 79.3%in the score of the joint index and CNS symptom positive group?double-positive group?,6.8%in the score of the joint index or CNS symptom positive group?single-positive group?,and 0 in the score of the joint index and CNS symptom negative group(double-negative group.?4?In the verification group,AUC of the score of the joint index was 0.897,and the sensitivity and specificity were 0.805 and 0.882,respectively.The mortality rates of double-positive,single-positive and double-negative groups in the verification group were 84.8%,25%and 2.19%,respectively.Compared with the disease outcomes of each subgroup in the verification group and the model group,the distribution of survival or death outcomes showed no statistically important difference between these two groups.?5?The overall analysis of the data of the model group and the verification group showed that the risk of death of SFTS patients treated with antibiotics decreased significantly,and there was a statistically significant difference between the treatment group and the untreated group?p<0.001?,but there were no significant differences in ribavirin and gamma globulin between the two subgroups?P value were 0.757 and 0.127?.Subgroup analysis showed that the risk of death in the single-positive group decreased significantly after treatment with ribavirin and antibiotics,and there was a significant difference between the treatment group and the untreated group regardless of patients with or without infection or agranulocytosis?P value were 0.045 and 0.003?;However,there was no significant difference between treated and untreated by ribavirin and antibiotic group in both double-positive group and double-negative group.Among the patients in each subgroup,there was no statistical difference between treated and untreated by gamma globulin group.Conclusion?1?According to the newly established model of SFTS according to the patients'scores of the joint index and CNS symptoms,the predictive incomes of the model group is consistent with that of the verification group,indicating that high prediction efficiency and good stability of this predictive model,which provides a practical method for early judgment of the prognosis of patients.?2?Ribavirin and antibiotics methods can only affect the disease outcomes of the patients in the single-positive group,but not in both double-positive group and double-negative group,also suggesting that the predictive model is stable and reliable and has a potential clinical guiding value for making the choices of SFTS treatment.