Looking beyond discharge: clinical variables at trauma admission predict long term survival in the older severely injured patient

Background: Long term follow up is difficult to obtain in most trauma settings, these data are essential for assessing outcomes in the older (>= 60) patient. We hypothesized that clinical data obtained during initial hospital stay could accurately predict long term survival. Study design: Using our trauma registry and hospital database, we reviewed all trauma admissions (age >= 60, ISS >15) to our Level 1 center over the most recent 7 years. Mechanism of injury, co-morbidities, ICU admission, and ultimate disposition were assessed for 2-7 years post-discharge. Primary outcome was defined as long term survival to the end of the last year of the study. Results: Of 342 patients discharged following initial admission, mean age was 76.2 +/- 9.7, and ISS was 21.5 +/- 6.9. 119 patients (34.8%) died (mean follow up 18.8 months; range 1.1-66.2 months). For 233 survivors, mean follow-up was 50.2 months (range 24.8-83.8 months). Univariate analysis disclosed post-discharge mortality was associated with age (80.1 +/- 9.64 vs. 74.2 +/- 9.07), mean number of co-morbidities (1.6 +/- 1.1 vs. 1.0 +/- 1.2), fall as a mechanism, lower GCS upon arrival (11.85 +/- 4.21 vs. 13.73 +/- 2.89), intubation at the scene and discharge to an assisted living facility (p < 0.001 for all). Cox regression analysis hazard ratio showed that independent predictors of mortality on long term follow-up included: older age, fall as mechanism, lower GCS at admission and discharge to assisted living facility (all = p < 0.0001). Conclusions: Nearly two-thirds of patients >= 60 who were severely injured survived >4 years following discharge; furthermore, admission data, including younger age, injury mechanism other than falls, higher GCS and home discharge predicted a favorable long term outcome. These findings suggest that common clinical data at initial admission can predict long term survival in the older trauma patient.