Although geriatric trauma patients are becoming more common, few large-scale analyses have comprehensively evaluated geriatric-specific characteristics in trauma. This study aims to clarify the age-specific characteristics, outcomes, and predictive accuracy of current trauma scoring systems among geriatric trauma patients.

Patients with severe trauma, with an Abbreviated Injury Scale score greater than or equal to 3, and registered in the Japan Trauma Data Bank during 2004 to 2015 were retrospectively reviewed. Age-related differences were assessed for injury mechanism, injured region, anatomic and physiologic severity, and inhospital mortality. The mortality risk was evaluated with multivariate mixed-effect models adjusted for Injury Severity Score, Revised Trauma Score, year of injury, and treating facility. Age-related differences in the accuracy of the Injury Severity Score and Revised Trauma Score for predicting inhospital mortality were evaluated with an area under the receiver operating characteristic curve.

We identified 127,303 patients, including 67,316 geriatric patients (52.9%) who were aged 60 years or older. The percentage of geriatric patients increased from 31.9% to 59.7% during the study period. The most frequent injury mechanism was ground-level falls (55.2%) and the most frequently injured region was the pelvis and lower extremities (43.7%). Severity-adjusted mixed-effects models revealed a marked age-dependent increase in mortality. Although the Injury Severity Score had similar predictive accuracy among all generations, the accuracy of the Revised Trauma Score decreased with increasing age.

The characteristics of trauma patients varied widely according to age, and mortality risk increased steadily with increasing age, despite a decrease in anatomic injury severity. The Revised Trauma Score had decreasing predictive accuracy at older ages, suggesting that an alternative measure is needed.