The relevance of 24 hour PRISM III score in predicting mortality in pediatric intensive care unit

Mortality is the most frequently assessed outcome in an ICU. The present study was designed to assess the relevance of 24 hour PRISM III score in predicting mortality in the PICU. This was a prospective analytic-descriptive study. 50 sick children consecutively admitted to PICU were enrolled after fulfilling the inclusion criteria and PRISM III score was calculated. The hospital outcome was recorded as died or survived. The predicted death was calculated. Formulae provided by the makers of PRISM III score, Statistical Package for Social Sciences SPSSTM version 15 and Microsoft Office Excel 2007 were used for calculating the PRISM III score and analysing the data. 50 children with mean age of 3.47 years (range 1-12); boys (58%) and girls (42%) were enrolled. Mean PRISM III score was 7.71±4.72 for survivors and 14.83±7.29 for non survivors. Overall mortality was 24%(12/50). 44%(22/50) had multiple associated disorders. 18 children had score of 0-5. The expected death in this group was 22.2% and the observed death was 5.56%. Among 16 children with the score of 6-10, expected mortality was 25% and observed was 18.75%. In 8 patients with the score of 11-15, expected mortality was 25.0% and observed mortality 37.5%. In 6 patients with score 16-20, observed death 50.0% and expected mortality was 16.67%. There was no significant difference between expected and observed mortality in any group (P<0.5). ROC analysis showed area under the curve of 82.6% .Cut off score of 15 had 67% accuracy. Mortality was significantly affected by the presence of multiple associated disorders and high PRISM III scores. Thus, PRISM III score has good predictive value in assessing the probability of mortality in children admitted to a PICU under Indian circumstances.