A new method for estimating probability of survival in pediatric patients using revised TRISS methodology based on age-adjusted weights

Background: TRISS methodology estimates probability of survival (PS) based on coefficients derived largely from adult data. We developed a novel pediatric age-specific method to estimate P-s. Methods: The Pennsylvania Trauma Outcome Study database was queried for pediatric patients injured between 1993 and 1996 (n = 9730). PS derived from the Pediatric Age-Adjusted TRISS (PART) methodology was generated using our Age-Specific Pediatric Trauma Score and Injury Severity Score with corresponding weights. A test data set of 7138 pediatric patients entered in the Pennsylvania Trauma Outcome Study from 1997 to 1999 was used to compute an expected number of survivors for PRAT, TRISS, and ASCOT (A Severity Characteristic of Trauma). Observed and expected survival were compared for blunt injured patients, for head injured patients, and by age category. Results: PAAT showed no significant difference between observed and expected survival. TRISS and ASCOT significantly underestimated overall survival: across age groups, for blunt injuries, for head injuries, and for patients whose PS was less than 91 %. Conclusion: PART offers a more reliable methodology than TRISS and ASCOT for comparing pediatric trauma outcomes.