DOI: 10.36347/sjams.2019.v07i04.016

Pages: 1447-1454

Objective: To compare the performance of recent versions of prognostic scores PRISM III (Pediatric risk of mortality) & PIM 3 (Pediatric index of mortality) scores at general pediatric intensive care unit, investigating the relation between observed mortality & survival & predicted mortality & survival. Methods: A prospective cohort study was undertaken at pediatric intensive care unit (PICU) of SVPPGIP, Cuttack during period of 1st July 2015 to 30th June 2016. Study was approved by institution’s Ethics committee. A total of 416 patients were enrolled out of 450 admissions during study period. Within first hour of admission PIM 3 was assessed & at 24 hours PRISM III was assessed. Patients were followed up for outcome measured in form of survivors & non survivors. Statistical analysis for model evaluation included Hosmer- Lemeshow goodness of fit test, receiver operating characteristics (ROC) curve & spearman’s correlation test. Results: A total of 416 patients were enrolled, 28 patients were excluded as their outcome was not known. Among the 388 patients 310 were survivors & 78 were no survivors. The overall number of estimated mortality was 65.86(16.97%) with PRISM III & 62.74% (16.17%) with PIM 3 compared to observed mortality of 78(20.1%).PIM 3(χ2= 3.71 P <0.05) & PRISM III (χ2= 2.23 p <0.05) had poor caliberation. PRISM III showed the better discrimination (ROC=0.893) followed by PIM 3 (ROC= 0.870). PIM 3 & PRISM III (0.851 to 0.927) revealed positive & significant correlation with spearman’s rank correlation (r= 0.318 P <0.0001). Conclusion: In this study PRISM III & PIM 3 under predicted mortality & also had poor caliberation with good discrimination. Overall both scores exhibited good capacity to discriminate between survivors & non survivors. They are tools with comparable performance at the prognostic evaluation of pediatric patients.