Objectives: To evaluate the relationship between mechanism of injury and mortality in geriatric trauma patients and the ability of existing injury severity indices (ISIs) to assess mortality.
Design: Retrospective review.
Setting: Urban level 1 trauma center.
Participants: Four thousand five hundred forty-five trauma patients age ≥55 presenting between 2008 and 2011.
Intervention: Low-energy (LE-GTP) and high-energy (HE-GTP) geriatric trauma patient cohorts were created based on ICD-9 injury codes. Existing ISIs were evaluated for their ability to predict in-hospital mortality using the area under the receiver-operating characteristic curve (AUROC).
Main outcome measures: Mortality.
Results: The Trauma Score-Injury Severity Score (TRISS) was the most predictive ISI for both cohorts and was deemed to have moderate predictive capacity (AUROC: 0.82) in LE-GTP and excellent predictive capacity (AUROC: 0.91) in the HE-GTP. For, HE-GTP each 1-year increase in age was associated with a 12% increase risk of mortality versus 6% for LE-GTP. Preexisting conditions (PECs) were distributed differently between the cohorts with significantly more PECs in the LE-GTP (P < 0.01).
Conclusions: Existing ISIs have fair-to-moderate predictive capacity for in-hospital morality in LE-GTPs and moderate-to-excellent predictive capacity in HE-GTPs. LE-GTPs and HE-GTPs are distinct cohorts that should be evaluated separately. Combining the cohorts underestimates both the effect of age on HE-GTPs and the effect of PECs on LE-GTPs while overestimating the effect of PECs on HE-GTPs.
Level of evidence: Prognostic Level II. See Instructions for Authors for a complete description of levels of evidence.