Article Text

Download PDFPDF

OP-079 Prediction of mortality using oxygen saturation (SpO2) in pediatric intensive care scores: validation of SpO2 modified PRISM-III score
  1. Enes Onkun1,
  2. Ebru Atike Ongun2,
  3. Hasan Serdar Kihtir2
  1. 1University of Health Sciences Antalya Training and Research Hospital, Department of Pediatrics, Antalya, Turkey
  2. 2University of Health Sciences Antalya Training and Research Hospital, Department of Pediatric Intensive Care, Antalya, Turkey


Aim To investigate the validation and discrimination of modified PRISM-III score using oxygen saturation (SpO2) in pediatric intensive care unit (PICU) admissions.

Material and Method A prospective observational study was conducted in PICU admissions between 1 month and 18 years at University of Health Sciences Antalya Training and Research Hospital between May-2021 and Aug-2022. Patient demographics, clinical variables within 48 hours and intensive care scores (PRISM-III, PIM-III, PELOD scores) were recorded. A modified PRISM-III score was developed by replacing SpO2 to PaO2 measurements. Group categorization was based on survival. SPSS-21 program was used for statistical analysis using Chi-square test for categorical variables and Mann-Whitney U test for continuous variables. ROC analysis calculated the cut-off value of the modified PRISM-III score to determine mortality. Kaplan Meier survival analysis determined the probability for survival, Cox regression model calculated the odds ratio for survival

Results Total of 380 patients were included (figure 1: age distribution). Common cause for PICU-admission was acute respiratory distress (47%). Mortality rate was 5.5%. The lowest SpO2 value measured within 24 hours was 90.3% (91.2% in survivors, 75.4% in the non-survivors group; p<0.05). PRISM-III, PIM-III, and PELOD scores were higher in the non-survivors group (p<0.05, p<0.05, p<0.05). The area under the curve for the modified PRISM-III score was 0.817 in ROC curve analysis, indicating that the discrimination was stronger than other mortality scores (figure 2). Kaplan-Meier and Cox regression analysis demonstrated a compatible outcome for SpO2-modified PRISM-III score on survival. As the modified PRISM-III score increased by one unit, the risk of mortality increased by 1.115 times (table 1).

Abstract OP-079 Figure 1

ROC analysis of intensive care scores.

Abstract OP-079 Table 1

Cox regression analysis for modifiye PRISM-III Score

Conclusions A novel SpO2-embedded mortality scoring model (SpO2-modified PRISM-III score) showed good predictive capability to determine mortality. A one-unit increase in the modified PRISM-III score causes a 1.115-fold increase in mortality. Comprehensive studies with larger patient populations are needed to develop easy-to-perform mortality scores in children.

  • pediatric intensive care
  • mortality score
  • PRISM score
  • oxygen saturation

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.