Background: D-dimer has been shown as a valuable predictor for the prognosis of sepsis. But the prognostic association of an elevated D-dimer with adverse outcomes of all critical illnesses in pediatric intensive care unit (PICU) has received far less emphasis.
Methods: This was a single-center retrospective study, including 7,648 critical patients aged between 28 days and 18 years from the pediatric intensive care (PIC) database from 2010 to 2018. The primary outcome was the in-hospital mortality rate.
Results: Higher levels of D-dimer, INR, PT, APTT, and lower Fib were observed in the non-survivor group (all P < 0.001). D-dimer, INR, PT and APTT were independent risk factors for prognosis in critically ill children. There was the highest AUROC in D-dimer for predicting in-hospital mortality of critically ill patients compared with INR, PT, APTT, and Fib (D-dimer: 0.77 vs. INR: 0.73 vs. PT: 0.73 vs. APTT: 0.64 vs. Fib: 0.60). The cut-off value, sensitivity, and specificity of D-dimer were 1.53, 0.65, and 0.77, respectively. Subgroup analysis showed a stable evaluation effectiveness of D-dimer for predicting in-hospital mortality of critically ill patients in the age and gender groups.
Conclusions: We found poorer coagulation function in the non-survivors compared with the survivors. Among the coagulation indicators, D-dimer was most strongly associated with in-hospital mortality of unselected critically ill children.
Keywords: coagulation dysfunction; critically ill children; in-hospital mortality; pediatric intensive care unit (PICU); plasma D-dimer.
Copyright © 2022 Wang, Liu, Xu and Liu.