TRISS methodology: an inappropriate tool for comparing outcomes between trauma centers

Area of Science:

• Trauma care research
• Medical outcomes analysis
• Statistical modeling in healthcare

Background:

• The Trauma and Injury Severity Score (TRISS) methodology is a standard for comparing survival outcomes between trauma centers.
• Previous studies have utilized TRISS for inter-center outcome comparisons.
• The current study aimed to assess TRISS's utility in comparing a large academic level I center with a small community level II center.

Purpose of the Study:

• To evaluate the role of TRISS in comparing survival outcomes between a large and a small trauma center.
• To assess the usefulness and limitations of TRISS across diverse patient subgroups.
• To determine if TRISS accurately reflects mortality differences between trauma centers of varying sizes and levels.

Main Methods:

• A trauma registry study comparing a level I and a level II trauma center.
• Utilized TRISS probability of survival, M value, and Z score for initial comparison.
• Employed matched-pair analysis and subgroup evaluation including sensitivity, specificity, and misclassification rates.

Main Results:

• The large center had a Z value of 2.24 (higher than expected mortality), while the small center had a Z value of -0.92 (lower than expected mortality).
• Matched-pair analysis showed similar Z scores (-0.40 and -0.95) for both centers, suggesting comparable outcomes.
• TRISS exhibited a high misclassification rate, exceeding 25% in some severe trauma subgroups.

Conclusions:

• The TRISS methodology demonstrates unreliability for comparing outcomes between trauma centers.
• TRISS has an unacceptably high misclassification rate, particularly in patients with severe trauma.
• The findings question the broad applicability of TRISS for inter-center outcome benchmarking.