Investigation on the risk of recurrent immune thrombocytopenia in children and establishment of a predictive model

Abstract

Objective To investigate the recurrence of immune thrombocytopenia (ITP) in children and to establish a predictive model. Methods A total of 288 children with ITP admitted to Children's Hospital of Wujiang District and Children's Hospital Affiliated to Suzhou University from January 2018 to April 2022 were collected. The factors potentially related to the recurrence of ITP in children were screened . The children in the model group were divided into 2 groups according to the presence or absence of recurrence and the indicators of the 2 groups were compared. After screening the potential influencing factors by LASSO regression and the independent influencing factors of relapse in children with ITP patients by Logstic regression analysis, we constructed a column-line graph model by using R language and validated it. Results A total of 37 (18.47%) of 201 patients in the model group experienced relapse. The age, blood type, duration of disease before treatment, antecedent infections, bleeding, initial treatment regimen, antinuclear antibody titer, initial count and mean platelet volume, initial platelet distribution width, initial peripheral blood lymphocyte count and time length to effective platelet count after treatment were found in the recurrence group versus the non-recurrence group The difference was statistically significant(P＜0.05).The results of multifactorial logistic regression analysis performed on the basis of LASSO regression showed that blood type, duration of illness before treatment, antecedent infection, initial treatment regimen, initial peripheral blood lymphocyte count, and time to effective platelet count after treatment were independent influences on the conversion of cough variant asthma to classic asthma in children. Based on the results of the multifactorial analysis, a column chart model for predicting ITP recurrence in children was developed in R. The results of the receiver operating characteristic(ROC) analysis showed that the area under curve(AUC) of the column chart model for predicting ITP recurrence in children in the modeling group was 0.867[95% CI(0.796, 0.938)] with a sensitivity of 84.2% and a specificity of 73.1%, and that in the validation group, the AUC was 0.838 [95% CI(0.765), 0.911] with a sensitivity of 82.3% and a specificity of 78.4%, 0.911)] sensitivity was 82.3% and specificity was 78.4%. The Bootstrap method was used to repeat the sampling 1 000 times, and the validation group was used for validation. The results of the calibration curve showed that the prediction curves of the model group and the validation group were basically fitted with the standard curve, suggesting that the model prediction accuracy was high. The results of the decision curve analysis of the model group showed that the net benefit rate of patients was greater than zero when the probability threshold of the column line graph model of predicting ITP recurrence in children was 0.15-0.75. Conclusions ITP recurrence in children is mainly affected by the patient's age, blood type, and pre-treatment course of the disease, and the column-line diagram model based on these factors has a high accuracy and differentiation for ITP recurrence in parenting children.

Keywords

• children|immune thrombocytopenia(itp)|multi factor analysis|nomogram