Construction And Verification Of A Clinical Prediction Rule For Severe Hand,Foot And Mouth Disease

Hand-foot-and-mouth disease(HFMD)is an acute infectious disease caused by enterovirus.Since its outbreak in China in 2008,its morbidity and mortality have increased year by year,and it has become one of the most important causes of child death in China.Many scholars have conducted researches concerning risk factor screening and forecast model construction to identify severe HFMD early in recent years.However,these studies remain in the qualitative evaluation stage,and couldn't be widely applied in clinic.In this study,children who were hospitalized in Children's Hospital Affiliation of Zhengzhou University from September 2016 to August 2018were recruited,severe HFMD clinical prediction Rules(CPRs)were constructed by using scoring model and nomogram,then validation and evaluation were performed on the rule.We constructed a simple and feasible CPR,and it's suitable in the clinical promoted application.Furthermore,the CPR for severe HFMD can help clinicians identify critical patients early and take timely treatment,which is the key to raise the cure rate,reduce the mortality and improve the prognosis.Objective1.Identify the practical predictors of severe HFMD.2.Construct clinical prediction rules for severe HFMD.MethodsChildren who were diagnosed with HFMD in the Children's Hospital Affiliation of Zhengzhou University from September 2016 to December 2017 were collected.Gradient lifting decision tree model and multivariable logistic regression were performed to screen the predictive factors,and construct scoring model and nomogram model,data collected in the same hospial from January 2018 to August2018 were used to verify the CPR.ResultsA total of 2154 cases of HFMD were included,with 1508 training samples and646 test samples.A total of 65 variables were included,and the results of multivariable logistic regression showed that,EV-71 infection(OR=2.304,95%CI:1.801~2.946),no less than 70%neutrophil ratio(OR=3.179,95%CI:1.658-9.550),Fever duration?2 days(OR=1.059,95%CI:1.032-1.086),White blood cell count?9.16×10~9/L(OR=1.124,95%CI:1.071-1.180)were the major risk factors.The 5 factors were used to construct scoring model and nomogram prediction model.The scoring model evaluation showed that the area under the receiver-operating characteristic curve(AUC)was 0.843,(P<0.05,95%CI(0.800-0.885)).The best cutoff value was 4 points,with a sensitivity of 80.4%,and a specificity of 75.1%.The nomogram evaluation showed that AUC was0.838(P<0.05,95%CI(0.795-0.880)).The best cutoff value was 62 points,the sensitivity was 78.2%,and the specificity was 72.9%.ConclusionsA simple and convenient CPR established by this study can perform quantitative prediction of severe HFMD,and can help the clinicians assess the risk of severe HFMD directly and accurately.