Values of procalcitonin and C-reactive protein on the mortalitiy risk prediction in critically ill pediatric patients

Purpose: To evaluate the value of procalcitonin and CRP levels in mortality rısk prediction and association with PIM, PRISM III in critically ill pediatric patients at Intensive Care Unit. Materials and Methods: In this singel-center study, we collected the datas of pediatric patients in the Pediatric Intensive Care Unit at Gaziantep University, Faculty of Medicine retrospectively. Including criteria consisted of children with 1 month to 18 years of age who has been hospitalized at least 48 hours in Pediatric Intensive Care Unit. Patients demographic features, clinical diagnoses, the presence of infection, physical examination findings, plasma Procalcitonin and C-reactive protein levels, laboratuary results, culture analyses and status (discharge or exitus..etc) were recorded. PIM and PRISM scores from the records of each cases were measured by SPSSwin statistical package version 20. CRP, Procalcitonin levels and PIM, PRISM results from exitus and alive patients compared statically. We identified the correlation between PRISM, PIM scores and CRP, procalcitonin levels; relationship between these results and the presence of infection, severity of disease. Results: Totally 47 pediatric patients ( female n:18; 38,3% of total, male n:29; 61,7% of total) were included in this study. Patients' mean age of 56.13± 61.9 months. We determined chronic diseases in 79% patients. During the follow up, approximately one third of cases died, the rest of them (63,8%) transferred to general pediatric service. There were infectious diseases in 59% of them; however septic shock caused by microorganisms was detected in half of these infected children (50%). PIM (p value: 0.013 ), PRISM (p value: 0.020 ) and levels of PCT in 0th, 24th, 48th hours (p values respectively: 0.000-0.017-0.008) were statistically significiant, higher in infected cases than the uninfected patients and despite of initial CRP level was higher in infected group, ıt was not statistically significant between two groups. In subgroup of infected patients with septic shock had increased CRP (p value: p=0.002 ) and PCT (p value: p=0.000), also these parameters were statistically significiant. According to analyses; there was no correlation between PIM II and initial levels of CRP, PCT (p value respectively: p=0.300, p=0.077). PCT level in 24th hours showed positive correlation, which was independet of infectious diseases, with PRISM (p=0,023). Lineer regression analysis for estimating PRISM score by the help of CRP and PCT 0th hour level indicated statistically non-significant (respectively p value: p=0.561, p=0.126), whereas we showed the correlation between PRISM and PCT level in 24th hours, regardless of infectious diseases. The PIM II and PRISM Scoring System were beneficial prediction methods for mortality (respectively p values: p:0.001, p:<0.001). To estimate the mortality risk, predictive value of PCT and CRP level at admittion date was found as low ( AUC: area under the curve; 0.684 and 569), when we compared with initial PIM score; predictive value of PCT 24 th and CRP 24 th hours levels was found low while we were compering with PRISM 24 th. Score (AUC: area under the curve; 0.601 and 0.485). Conclusion: We would like to report that PCT is correlated with severity of disease and mortality risk in critically ill pediatric patient at ICU, also ıt could be important biochemical marker to determine the prognosis. By the way, we suppose that procalcitonin may be used for measuring the PRISM score and estimate the prognosis and mortality risk as well as PIM II, PRISM in Pediatric Intensive Care Units. Our limitaions was single-center, retrospective study and small number of cases. This research requires multicenter-multidisplinary studies about PCT effects on determination of mortality risk eywords: C-reactive protein (CRP), Procalcitonin (PCT), PIM II, PRISM