Clinical And Epidemiological Characteristics Of Hand-foot And Mouth Disease And Derivation Of A Mortality Risk Score For Severe Hand,Foot And Mouth Disease

Objectives(1)To analyze the clinical and epidemiological charateristics including population distribution,seasonal distribution,the composition of enterovirus pathogen,the complications and the case fatality rate,and to explore the composition of pathogen in HFMD and SHFMD,then provide the guide for prevention and control of HFMD.(2)To screen laboratory indicators related with the severity of illness and complication,and evaluate the performance of the selected indicators clinical outcomes and prognosis of SHFMD.The results can provide a reference for develop a severity or mortality risk score of severe hand,foot and mouth disease.(3)To develop a mortality risk score(MRSHFMD)that is based on laboratory indicators which predict the complications outcomes for children with severe HFMD,and compare its validity with PCIS and PRISM.Methods(1)Part 1:There were 7 203 hospitalized patients who diagnosed with HFMD between 1 January,2013 and 31 December,2017 in Hunan Children's Hospital were included for the analysis.The data such as basic information,complications and outcome were obtained from medical records.(2)Part 2:All of 866 SHFMD included in the present study were from 7 203patients with HFMD treated in Hunan Children's Hospital between January 2013 and December 2017.Ten laboratory indicators were evaluated by R clustering method in system cluster and area under the receiver operating characteristic curve(AUC)regard to the performance of outcomes and prognosis in SHFMD patients.(3)Part 3:All of 866 SHFMD enrolled in the present study were from 7 203patients with HFMD treated in Hunan Children's Hospital between January 2013 and December 2017.866 with SHFMD patients were divided into derivation and validation group according to number of single and double.Multivariate logistic regression analysis was used to develop a preliminary model.A simple integer-based point score for each predictor variable was established by dividing beta coefcients by the absolute value of the smallest coefcient in the model and round up to the nearest integer.Discrimination of the mortality risk score between death and survival was assessed by calculating the area under the receiver-operating characteristic curve(AUC)in the derivation and validation group.Calibration of the mortality risk score was evaluated using the Hosmer–Lemeshow goodness-of-fit?~2 test in the derivation and validation group.Results(1)The distribution of age:The majority of HFMD patients(6113/7203,84.87%)were younger than 3 years old group.The number of SHFMD patients(420/3209,13.09%)were the highest in 2 years old group,but the number of SHFMD patients(153/1617,9.46%)were the lowest in 1 years old group.The case fatality rate(20/1278,1.55%)was the highest in 3 years old group.The case fatality rate(3/1090,0.28%)was the lowest in>3 years old group.The percentage of patients with SHFMD in>2 years old group was significant higher than that of in?1 years old group(?~2=17.92,P<0.001).The rate of complications was significant higher in cases of>2 years old than that of cases of?2 years old(?~2=164.61,P<0.001).The highest case-fatality rate(1.55%)was found in 3 years old groups,which was significantly higher than that of in other age groups(?~2=14.31,P=0.003).The predominant pathogen was other enterovirus and EV-A71,and the mixed cases with two enterovirus were common in over 2 years old group.The predominant pathogen was other enterovirus in patients with?1 year old group.(2)The distribution of years.Between the year of 2014~2017,more HFMD cases(1931)were hopitalized in the year of 2014 than that of other years.The proportion of SHFMD patients(351/1931,18.18%)and the case fatality rate(33/1931,1.71%)were the highest in the year of 2014.The number of HFMD patients was the fewest,and1126 cases were observed in 2017.The proportion of SHFMD patients(59/1126,5.24%)was the lowest in 2017,and no dead case was observed in 2017.The proportion of cases with SHFMD in 2014 were significant higher than that of in other years(?~2=157.11,P<0.001).The peak rate of complications(56.46%)was observed in2016,the peak case fatality rate(1.71%)was observed in 2014,and the incidence of complications or the case fatality rate was statistically significant different from 2013to 2017(P<0.001).The proportion of EV-A71 patients in a whole year showed a fluctuate tendency between 2013 and 2017.The proportion of CV-A16 patients(261/1520,17.17%)and mixed with two EVs patients(384/1520,25.26%)was the highest in 2015.The proportion of other enterovirus patients was the highest in 2017.(3)The distribution in type of enterovirus.Most HDFD cases were happened during May to July and Spetember to October The EV-A71were more likely to be dectected between April and July,while CV-A16 were more often to be detected between April and June.The proportion of SHFMD infected with EV-A71(23.88%)was higher than that of SHFMD infected with other enterovirus(?~2=317.23,P<0.001).Among SHFMD patients,EV-A71,other EV and mixed with EV-A71 and CV-A16,accounted for 38.04%,29.92%and 21.69%respectively.The rate of complications or the case fatality rate was 58.24%or 1.89%for EV-A71.The rate of complications or the case fatality rate was significant different in patients with different pathogens(P<0.001).(4)Stage II patients had the lowest NT-proBNP,WBC,GLU,LAC,CK and Mb level and stage IV patients had the highest NT-proBNP,WBC,GLU,LAC,CK and Mb level,and the difference between the groups was statistically significant(P<0.05).The level of NT-proBNP,WBC,GLU,LAC,CK and Mb was increased with the severity of HFMD.The level of NT-proBNP,WBC,GLU,LAC,CK and Mb was higher in the patients with brainstem encephalitis,pulmonary edema,pulmonary hemorrhage,or circulatory failure than that of without complications or death(P<0.001).(5)The area under the receiver operating characteristic(ROC)curve and its95%CI were 0.675(0.623~0.726),0.811(0.750~0.871),0.874(0.821~0.935),0.776(0.703~0.930),and 0.897(0.842~0.951)to discriminate between patients with and without brainstem encephalitis,pulmonary edema,pulmonary hemorrhage,circulatory failure,and death,respectively,using NT-proBNP.The area under the receiver operating characteristic(ROC)curve and its 95%CI were 0.613(0.560~0.665),0.750(0.683~0.818),0.827(0.762~0.891),0.714(0.646~0.781)and0.865(0.806~0.925)to discriminate between patients with and without brainstem encephalitis,pulmonary edema,pulmonary hemorrhage,circulatory failure,and death,respectively,using WBC.The area under the receiver operating characteristic(ROC)curve and its 95%CI were 0.668(0.619~0.718),0.771(0.706~0.836),0.839(0.771~0.908),0.767(0.704~0.829),and 0.855(0.781~0.929)to discriminate between patients with and without brainstem encephalitis,pulmonary edema,pulmonary hemorrhage,circulatory failure,and death,respectively,using GLU.The area under the receiver operating characteristic(ROC)curve and its 95%CI were0.618(0.564~0.672),0.723(0.648~0.798),0.801(0.721~0.881),0.744(0.677~0.812)and 0.829(0.743~0.916)to discriminate between patients with and without brainstem encephalitis,pulmonary edema,pulmonary hemorrhage,circulatory failure,and death,respectively,using LAC.The area under the receiver operating characteristic(ROC)curve and its 95%CI were 0.579(0.526~0.632),0.721(0.656~0.785),0.777(0.709~0.845),0.641(0.570~0.713),and 0.811(0.740~0.882)to discriminate between patients with and without brainstem encephalitis,pulmonary edema,pulmonary hemorrhage,circulatory failure,and death,respectively,using CK.The area under the receiver operating characteristic(ROC)curve and its 95%CI were 0.659(0.608~0.709),0.719(0.646~0.785),0.808(0.737~0.879),0.718(0.655~0.781),0.807(0.723~0.891)to discriminate between patients with and without brainstem encephalitis,pulmonary edema,pulmonary hemorrhage,circulatory failure,and death,respectively,using Mb.(6)On the basis of Kaplan–Meier method,the case fatality rate in SHFMD with higher the level of NT-proBNP,WBC,GLU,LAC,CK and Mb was 41.41%,30.19%,24.14%,18.35%,20.96%and 22.84%.And the case fatality rate in SHFMD with lower the level of NT-proBNP,WBC,GLU,LAC,CK and Mb was 1.96%,3.16%,2.91%,4.76%,3.00%and 2.70%.(7)The results of multivariate logistic regression showed that NT-proBNP,GLU,and WBC were significantly associated with non-survivors of severe HFMD.The?value of the blood glucose was 1.83,with a score of 2 as the reference.The mortality risk score,comprising WBC counts,GLU,and NT-proBNP,namely GLU>7.8mmol/L,WBC>16.5×10~9/L,NT-proBNP 1.3~10 ng/mL,and NT-proBNP>10 ng/mL was given 2,2,2 and 3 points value.The range of score is from 0 to 7 points.(8)The AUCs(95%CI)was 0.90(95%CI:0.82~0.98)in the derivation group.Using the“optimal”cutoff point of 3,the sensitivity,specificity,positive predictive value,negative predictive value and Youden's index were 75.00%,94.34%,46.67%,98.28%and 69.34%in the derivation group,respectively.The AUCs(95%CI)was0.94(95%CI:0.89~0.99)in the validation group,respectively.Using the“optimal”cutoff point of 3,the sensitivity,specificity,positive predictive value,negative predictive value and Youden's Index were 71.43%,95.08%,51.28%,97.87%and66.51%in the validation group,respectively.The AUC for the mortality risk score showed no significant difference between derivation and validation groups tested by the C-statistic(Z=-0.91,P>0.05).(9)The AUCs(95%CI)were 0.83(0.75~0.92)for PCIS,0.87(0.80~0.95)for PRISM,0.93(0.90~0.96)for MRSHFMD.The AUC between the MRSHFMD and PCIS showed a significant difference by the C-statistic(P<0.05).The AUC between the MRSHFMD and PRISM showed no significant difference by the C-statistic(P>0.05).Conclusions(1)HFMD is mainly caused by other enterovirus and enterovirus 71(EV-A71),and the marjority of HFMD patients were younger than 3 years old.For the number of HFMD patients,the peak appeared between April and June whilst the minor peak observed between Spetember and October.The number of HFMD patients were increased between 2013 and 2014,and then decreased between 2014 and 2017.(2)From the point view as the proportion of enterovirus,other enterovirus has displaced EV-A71 and CV-A16 to become the predominant serotype in 2017.EV-A71infections accounted for the majority of children with severe HFMD.Non-SHFMD is mainly caused by other untyped enteroviruses.The porportion of complications or the fatality was the highest in patients infected with EV-A71.(3)The level of NT-proBNP,WBC,GLU,LAC,CK and Mb reflect the the severity of disease in patients with HFMD.(4)The level of NT-proBNP,WBC,GLU,LAC,CK and Mb are suitable to predict well in pulmonary edema/pulmonary hemorrhage/circulatory failure and mortality in children with HFMD,but not to predict brainstem encephalitis in children with HFMD.The NT-proBNP was a better biomarker than other indicators to discriminate HFMD patients with or without severe complications or death.(5)The MRSHFMD was comprised three readily available laboratory indicators(WBC,GLU,and NT-proBNP).Compared with PCIS or PRISM,MRSHFMD has less indicators and more accurate,which could be used to predict the mortality risk of SHFMD in time,and it is more suitable to be used in primary hospital with insufficient equipment or health professionals.