This retrospective study included pediatric patients admitted to the PICU of The First Affiliated Hospital of Bengbu Medical College, from April 2015 to December 2019 who took both blood routine examination and coagulation function test within 24 hours of admission. Patients below 28 days or above 18 years of age, Patients whose blood routine examination or coagulation function test exceeds the detection range of the instrument cannot be statistically analyzed, Patients with diseases that cause coagulation to disorder such as rodenticide poisoning, and patients with information incomplete were excluded from this study. Patients who have at least five of the following ten physiological indicators (Heart rate, blood pressure (systolic blood pressure), respiratory rate, oxygen partial pressure, blood pH, blood sodium, blood potassium, creatinine or urea nitrogen, hemoglobin level, gastrointestinal system performance) to complete the PCIS score were included in this study.
Clinical and laboratory data
Data on patients’ age, sex, duration of hospitalization, admission diagnosis, and discharge status were collected. Both blood routine examination and coagulation function test are determined by Japanese Sysmex-XN900 and Sysmex-CS5100. The following admission data were also collected: leukocyte (WBC), erythrocyte (RBC), hemoglobin (HB), Hematocrit (HCT), mean red blood cell volume (MCV), mean hemoglobin (MCH), mean hemoglobin concentration (MCHC), red blood cell distribution width-CV value (RDWCV), red blood cell distribution width-SD value (RDWSD), Neutrophils (N), lymphocytes (L), monocytes (MONO), Platelet (PLT), Mean platelet volume (MPV), plateletcrit (PCT), Platelet distribution width (PDW), platelet-large cell ratio (P-LCR), Prothrombin time (PT), international standardized ratio (INR), prothrombin activity (PA), thrombin (TT), activated partial thrombin time (APTT), fibrinogen (FIB), and D-dimer (DD). PT prolongation is calculated by PT subtract control.
We used the worst indicators within 24 hours of the patients’ admission for PCIS score, with scores >90% of the total score as non-critical, and ≤ 90% of the total score as critical .
Based on the PCIS score, each patient was subdivided into two groups: non-critical and critical. SPSS Statistics 25 software was used for statistical analysis. R software with RMS packages was used for the nomogram compute. Continuous variables were presented as median and interquartile due to significant skewness. In the comparison of variables distributed homogeneously, the t-test was used for continuous variables. For variables not showing homogeneous distribution, the Mann-Whitney U-test was used. P<0.05 was accepted as statistically significance. Univariate logistic regression analysis was performed to identify risk factors for critical illness. All P values were two-sides and risk factors with P values <0.05 in univariate analysis were included in a multivariate analysis. Multivariate logistic regression analysis was performed to identify independent risk factors, and a stepwise method was used to identify the useful combination of factors that could most precisely predict critical illness. Using the R language, a nomogram based on multivariable logistic analysis was established to predict the critical of pediatric illness. The predictive performance of this model was then evaluated using the concordance index(C-index). The receiver operating characteristic (ROC) curves were used to evaluate and compare the abilities for predicting pediatric critical illness.