The accurate prediction of the prognosis for critically ill children is crucial, with the Pediatric Index of Mortality (PIM) and Pediatric Risk of Mortality (PRISM) being extensively utilized for this purpose. Albumin, C-reactive protein (CRP), and lactate levels, which are indicative of inflammation and circulatory status in critically ill children, have not been incorporated into existing scoring systems. This retrospective cohort study evaluated the association between biological markers and the clinical outcomes in children with critical illnesses. PRISM-4 and PIM-3 death probability (DP), albumin, lactate, CRP, and CRP/albumin ratio were recorded upon admission. The accuracy of the indexes in predicting mortality were assessed by calculating the area under the curve (AUC). There were 942 patients included and the 28-day mortality rate was 7.9%. The AUC for PRISM-4, PIM-3, CRP, CRP/albumin ratio, albumin, and lactate were 0.923, 0.896, 0.798, 0.795, 0.751, 0.728, respectively. The findings in the subgroup analysis of septic patients were similar to those found in the overall population. Although CRP, CRP/albumin ratio, albumin, and lactate levels are all linked to mortality in children, CRP and the CRP/albumin ratio have lower predictive values than albumin and lactate. Incorporation of albumin and lactate into scoring systems will improve predictability.
Keywords: albumin; lactate; mortality; pediatric critical care; sepsis.