Related Experiment Video
Updated: Jun 22, 2026
Inverse Probability of Treatment Weighting (Propensity Score) using the Military Health System Data Repository and National Death Index
Published on: January 8, 2020
The impact of injury coding schemes on predicting hospital mortality after pediatric injury
Randall S Burd1, David Madigan
1Division of Trauma and Burns, Children's National Medical Center, Washington, DC, USA. rburd@cnmc.org
Insights
Combining injury coding schemes like ICD-9, AIS, and Barell matrix improves pediatric trauma mortality prediction. This approach enhances model accuracy for better patient outcomes.
Area of Science:
- Trauma research
- Pediatric critical care
- Health informatics
Background:
- Accurate adjustment for injury severity is crucial for evaluating trauma management effectiveness.
- The choice of injury coding scheme impacts predictive model performance.
- The combined effect of different injury coding schemes on model performance is not well-understood.
Purpose of the Study:
- To develop Bayesian logistic regression models for predicting hospital mortality in injured children.
- To compare the performance of models using individual and combined injury coding schemes.
Main Methods:
- Utilized data from the National Trauma Data Bank (NTDB) and National Pediatric Trauma Registry (NPTR).
- Developed Bayesian logistic regression models using International Classification of Disease-9th revision (ICD-9) codes, Abbreviated Injury Scale (AIS) scores, and the Barell matrix, individually and in combination.
- Evaluated model performance on independent datasets from NTDB and the Kids' Inpatient Database (KID).
Main Results:
- Models incorporating both ICD-9 and AIS codes demonstrated optimal discrimination (AUC = 0.921-0.967).
- Models using the Barell matrix showed optimal calibration (HL h-statistic = 19-69).
- Combined coding schemes improved discrimination and calibration compared to ICD-9 alone.
Conclusions:
- Incorporating multiple injury coding schemes enhances mortality prediction models in pediatric trauma.
- Combining coding schemes offers a promising strategy for improving the accuracy of injury mortality estimates.
Objectives:
Accurate adjustment for injury severity is needed to evaluate the effectiveness of trauma management. While the choice of injury coding scheme used for modeling affects performance, the impact of combining coding schemes on performance has not been evaluated. The purpose of this study was to use Bayesian logistic regression to develop models predicting hospital mortality in injured children and to compare the performance of models developed using different injury coding schemes.
Methods:
Records of children (age < 15 years) admitted after injury were obtained from the National Trauma Data Bank (NTDB) and the National Pediatric Trauma Registry (NPTR) and used to train Bayesian logistic regression models predicting mortality using three injury coding schemes (International Classification of Disease-9th revision [ICD-9] injury codes, the Abbreviated Injury Scale [AIS] severity scores, and the Barell matrix) and their combinations. Model performance was evaluated using independent data from the NTDB and the Kids' Inpatient Database 2003 (KID).
Results:
Discrimination was optimal when modeling both ICD-9 and AIS severity codes (area under the receiver operating curve [AUC] = 0.921 [NTDB] and 0.967 [KID], Hosmer-Lemeshow [HL] h-statistic = 115 [NTDB] and 147 [KID]), while calibration was optimal when modeling coding based on the Barell matrix (AUC = 0.882 [NTDB] and 0.936 [KID], HL h-statistic = 19 [NTDB] and 69 [KID]). When compared to models based on ICD-9 codes alone, models that also included AIS severity scores and coding from the Barell matrix showed improved discrimination and calibration.
Conclusions:
Mortality models that incorporate additional injury coding schemes perform better than those based on ICD-9 codes alone in the setting of pediatric trauma. Combining injury coding schemes may be an effective approach for improving the predictive performance of empirically derived estimates of injury mortality.
