Construction Of Risk Prediction Model For Urinary Retention Of PICU Hospitalized Children During Mechanical Ventilation

Objective:By studying the actual occurrence of urinary retention in children in the Pediatric Intensive Care Unit PICU after mechanical ventilation,analyzing the characteristics of relevant risk factors,screening out independent risk factors,and constructing and validating a risk prediction model for urinary retention in mechanically ventilated children,we aim to provide an effective evaluation tool for enhancing the management of children in the PICU,reducing the incidence of urinary retention,and improving the prognosis.The aim was to provide a valid assessment tool to enhance the management of PICU children,reduce the incidence of urinary retention and improve the prognosis.Methods:A retrospective survey study was used for the modeling group,and children attending the PICU of Jiangxi Children’s Hospital from January 2019 to December2020 who met the nadir criteria were selected for the study.The included children were renumbered according to the order of hospitalization number and 190 cases were selected for modeling using random sampling method.Risk factors related to the occurrence of urinary retention in mechanically ventilated children were collected by reviewing medical records,and subject operating characteristic curve(ROC)analysis was performed using Medcal software according to the presence or absence of urinary retention in children,and continuous variables were transformed into dichotomous variables according to the best cut-off values.The risk factors for the occurrence of urinary retention in mechanically ventilated children were initially screened by univariate analysis,and the factors,especially those with statistically significant differences in the results of univariate analysis,were included in the binary logistic regression analysis to establish a logistic regression model for the occurrence of urinary retention in mechanically ventilated children.To facilitate clinical use,the risk factors in the logistic regression model were assigned scores according to their bias regression coefficients,and "rms","Hmisc",and "survival" were used in the R 3.5.2 software.The model performance was evaluated by using the consistency index(C-index)and calibration curves.The model group adopted a prospective study.From September 2021,the children admitted to PICU of Jiangxi Children’s Hospital were collected.By April 2022,a total of 95 patients who met the criteria for inclusion and excretion of the model group were collected.The established risk prediction model for children with mechanical ventilation was validated with the data of the model group.Result:1.In this study,285 mechanically ventilated children in PICU who met the discharge criteria were collected,and the incidence of urinary retention in mechanically ventilated children in the modeling group was 71.05%(135/190);the incidence of urinary retention in mechanically ventilated children in the test group was 69.47%(66/95).2.The results of ROC curve analysis showed that AUC of age,body index,length of hospitalization and duration of mechanical ventilation were 0.761,0.556,0.602 and 0.649 respectively;The best cutoff values were 8.6 months,22.6 kg/m2,12 days and 131 hours,respectively.3.In univariate analysis,the proportion of mechanically ventilated children aged >8.6 months,the proportion of APACHE II score ≥ 20,the proportion of hospitalization days >12 days,the proportion of mechanical ventilation time >131 h,the proportion of sedative drugs used,the proportion of analgesic drugs used,and the proportion of inotropic drugs used were all significantly higher in the mechanically ventilated children with urinary retention group than in the mechanically ventilated children without urinary retention group,and the differences were statistically significant(P < 0.05).4.Multifactorial logistic regression analysis showed that days of hospitalization >12 days(OR=1.045,95% CI: 1.034-1.061,P=0.014),duration of mechanical ventilation(OR=1.072,95% CI: 1.050-1.093,P=0.002),use of sedative medication(OR=1.016,95% CI.1.012-1.019,P=0.011),use of analgesic medication(OR=1.107,95% CI: 1.065-1.151,P=0.005),and use of inotropic medication(OR=1.128,95% CI: 1.054-1.206,P=0.009)were risk factors for the development of urinary retention in mechanically ventilated children.5.The risk prediction model constructed with independent risk factors such as days of hospitalization,duration of mechanical ventilation,use of sedative drugs,use of analgesic drugs,and use of inotropic drugs as predictors had good predictive accuracy.The internal validation results showed that the C-index of the risk prediction model for predicting urinary retention in mechanically ventilated children was 0.762(95% CI: 0.681-0.846).The calibration curve showed good agreement between the observed and predicted values.The risk prediction model predicted urinary retention in mechanically ventilated children with a threshold >0.12,and the risk prediction model provided a net clinical benefit;in addition,the risk prediction model provided a net clinical benefit over the number of days in hospital,duration of mechanical ventilation,use of sedative medication,use of analgesic medication,and use of inotropic medication.6.The variables of the test model group were brought into the established prediction model of the risk of urinary retention in mechanically ventilated children.The ROC curve method and the H-L goodness-of-fit test were used to test the differentiation and consistency of the prediction of the risk of urinary retention in mechanically ventilated children.The area under the curve AUC was 0.823(95% CI:0.732-0.884),2 = 10.275,p = 0.246(p > 0.05)for the H-L goodness-of-fit test,and The area under the curve AUC was 0.823(95% CI: 0.732-0.884),2 = 10.275 for the H-L goodness-of-fit test,p = 0.246(p > 0.05).It is suggested that the established model for predicting the risk of urinary retention in mechanically ventilated children has good discrimination and consistency.The specificity of the prediction model was68.97%,the sensitivity was 89.39%,the negative predictive value was 67.74%,the positive predictive value was 87.50%,and the overall prediction accuracy was81.05%.Conclusion:1.Longer hospital stay(>12 days),longer duration of mechanical ventilation(>131h),use of sedative drugs,use of analgesic drugs,and use of inotropic drugs were independent risk factors for the development of urinary retention in mechanically ventilated children.2.In this study,a risk prediction model for urinary retention in mechanically ventilated children was constructed based on the above risk factors,and the model was proven to improve the accuracy of predicting urinary retention in mechanically ventilated children with good clinical efficacy,which can provide some reference basis for clinical intervention.