Clinical Course And Prediction Model Of Death Risk In Severe Fever With Thrombocytopenia Syndrome

[Background]Severe fever with thrombocytopenia syndrome(SFTS)is a new infectious disease caused by SFTSV.SFTSV is an RNA virus,which belongs to the Buniavirus.It can be transmitted through tick bite or through contacting with body fluids of a sick person.Patients with SFTS usually present with high fever,thrombocytopenia,leukopenia,with respiratory,gastrointestinal and neurological symptoms,severe case can progress to multiple organ dysfunction.At present,the case fatality rate is reported between 5.3%and 47%.The incidence of SFTS is increasing year by year and the case fatality rate is always high.Moreover,there is no effective vaccine or specific treatment for SFTS.SFTS has become a new infectious disease that seriously threatens public health.The clinical course,risk factors and early prediction for death risk are unknown.[Objective]Through retrospective analysis of the epidemiology and clinical characteristics of patients with SFTS,our study aims to analyze the clinical course of SFTS and the risk factors related to death.The prediction model of death risk SFTS was established to guide clinical practice.[Method]Collecting the confirmed SFTS patients from the Jinan Infectious Diseases Hospital from January 1,2011 to December 30,2018.The diagnostic criteria are based on the"Guidelines for the Prevention and Treatment of Severe Fever with Thrombocytopenia Syndrome(2010 Edition)" issued by the China Ministry of Health.Demographic and clinical data were collected,patients were divided into recovered group and death group.The categorical variables were expressed as percentages,and the chi-square test was used;continuous variables were expressed as mean ± standard deviation(SD)or median(minimum-maximum).Variables that fit the normal distribution were tested using the t test;variables that did not fit the normal distribution were tested using the Mann-Whitney test.Multivariate Logistic regression analysis was performed using variables with a P<0.05 in univariate analysis to determine risk factors associated with SFTS death.An ideal risk prediction model was formed through regression equation and considering specific clinical practice and simplicity,risk factors in the model are assigned to determine the scoring criteria and the scoring risk level is verified.[Result]On January 2011 and November 2018,a total of 694 laboratory-confirmed SFTS patients’ medical reports were collected,of which 145 died,with a case fatality rate of 20.9%(95%CI:17.6%-23.9%).329 cases were male and 365 cases were female;The average age was 61.65±11.27years(range:15-89years);The incidence was highest from May to October,accounting for 98.4%;568 patients(81.8%)were farmers.The incubation period of SFTS in the study samples was about 5 days(IQR:2-16).Patients usually present with fever and anorexia,accompanied by the gastrointestinal,hematological,respiratory and neurological symptoms.The results of laboratory tests showed a decrease in platelet counts,and some patients may have abnormal laboratory results in liver function,renal function and coagulation function.Through the analysis of heart rate and body temperature of patients,we found that relative bradycardia was one of the clinical manifestations of SFTS.The risk of death was significantly higher in patients≥60 years of age than in patients<60 years of age(P<0.001,OR=3.42,95%CI:2.20-5.31).There was no significant difference death risk between men and women(P>0.05,23.7%vs.18.4%).Risk of death of patients with drinking history were 2.1 times than those without drinking history(P<0.05,OR=2.1,95%CI:1.30-3.38).However,there was no significant difference in the risk of death between patients with or without smoking history,tick bite history,diabetes mellitus,hypertension and coronary heart disease(P>0.05).There was no significant difference in the risk of death between patients living in urban areas and rural areas,or within and outside Jinan(P>0.05).Univariate analysis of risk factors of death revealed that the following symptoms and signs and laboratory examination results were significantly different between the recovered group and death group:hemorrhagic signs(OR=4.62,95%CI:3.12-6.87),hematemesis(OR=8.90,95%CI:4.62-17.13),melena(OR=2.57,95%CI:1.67-3.98),epistaxis(OR=10.63,95%CI:2.78-40.59),oral or giving bleeding(OR=6.48,95%CI:3.65-11.49),dyspnea(OR=65.87,95%CI:37.90-114.5),pulmonary infection(OR=3.29,95%CI:2.04-5.30),neurological symptoms(OR=62.42,95%CI:25.90-91.08),ecchymosis(OR=4.69,95%CI:3.16-6.98),ecchymosis at venipuncture site(OR=4.3,95%CI:2.73-6,76),disturbance of consciousness(OR=48.57,95%CI:25.90-91.08).N%(OR=3.26,95%CI:2.21-4.83),L%(OR=3.03,95%CI:2.07-4.46),PLT(OR=2.14,95%CI:1.04-4.39),INR(OR=2.81,95%CI:1.83-4.32),APTT(OR=8.18,95%CI:4.95-13.54),TT(OR=2.92,95%CI:1.78-4.79),Dd(OR=3.34,95%CI:2.11-5.30),ALT(OR=2.26,95%CI:1.37-3.72),AST(OR=16.78,95%CI:2.30-122),TB(OR=3.62,95%CI:1.93-6.76),BUN(OR=6.22,95%CI:3.79-10.22),Cr(OR=9.14,95%CI:4.63-17.93),CK(OR=5.86,95%CI:3.00-11.45),CKMB(OR=5.05,95%CI:336-7.60),LDH(OR=4.43,95%CI:2.72-7.23).Multiple collinearity test and multivariate Logistic regression analysis showed that independent risk factors associated with death included:age≥60years(OR=3.55,95%CI:1.50-8.39),neurological symptoms(OR=16.30,95%CI:5.26-50.53),dyspnea(OR=16.15,95%CI:7.88-33.11),INR>1.2(OR=4.66,95%CI:1.90-11.43),APTT>42s(OR=6.50,95%CI:2.95-14.30),TB>25μmol/L(OR=4.84,95%CI:1.41-16.64).Through the Logistic regression equation and considering specific clinical practice and simplicity,the prediction model was finally formed as follows:P=ea/(1+ea),a=-6.944+1.267×age ≥60 years(0 or 1)+2.791×neurological symptoms(0 or 1)+2.782×dyspnea(0 or 1)+1.539×INR>1.2mmol/L(0 or 1)+1.872×APTT>42s(0 or 1)+1.576×TB>25μmol/L(0 or 1)(e was an exponential function and P was the probability of death in hospital for SFTS patients).The Hosmer-Lemeshow goodness-of-fit test of the model showed that P=0.913>0.05,which proved that the model had a good fitting effect ROC curve showed that the area under the curve A=0.956(P<0.001,95%CI:0.939-0.973),the sensitivity was 91%and the specificity was 86.8%.We built a scoring criteria according to each of the risk factors in the established risk prediction model:age≥60 years:lpoint,neurological symptoms:3 points,dyspnea:3 points,INR>1.2:2 points,APTT>42s:2 points and TB>25μmol/L:2 points.The lowest score was 0,the highest score was 13.The low risk was 0-3 points,medium risk was 4-8 points and high risk was 9-13 points.According to the risk rating of the cases in our cases,the CRF in the low-risk,medium-risk and high-risk group were 0.65%,20.39%and 91.76%,respectively.The prediction model of death risk and the death risk rating score were well.[Conclusion]SFTS caused by SFTSV infection has complicated clinical manifestations involving multiple systems and organs.Relative bradycardia was one of the clinical manifestations.Drinking history was one of the risk factors for death.The prediction model of death risk,which established according to the independent risk factors for death(age≥60 years,neurological symptoms,dyspnea,INR>1.2,APTT>42s,TB>25μmol/L),has a good matching effect and can be applied to clinical practice.