Does Hospital Experience Rather than Volume Improve Outcomes in Geriatric Trauma Patients?

Background: Although high absolute hospital geriatric trauma volume (GTV) is associated with improved outcomes among geriatric trauma patients, the actual geriatric trauma proportion (GTP) might be a better predictor of outcomes. Study Design: Adult trauma admissions were identified in the California State Inpatient Database, 2007 to 2011. Hospital characteristics were extracted from the American Hospital Association database. The annual trauma volume of patients 65 years and older was calculated for each hospital. The GTP was derived by dividing the GTV by the overall adult trauma volume and hospitals were categorized into tertiles of GTP. Outcomes were hospital mortality, failure to rescue (FTR), and 30-day readmission in geriatric trauma patients. Independent risk factors were assessed with clustered multivariate logistic regression models adjusted for patient and hospital characteristics. Results: There were 61,915 geriatric trauma patients included from 63 trauma centers. Hospital mortality, FTR, and 30-day readmission rates were 4.99%, 16.07%, and 12.03%, respectively. The adjusted odds ratios and 95% CIs for in-hospital mortality and FTR per 100 patient increase in GTV were 0.91 (95% CI, 0.83-1.00) and 1.01 (95% CI, 0.90-1.14), respectively. As compared with hospitals in the lowest tertile, adjusted odds of mortality and FTR in the highest tertile were 0.71 (95% CI, 0.54-0.94) and 0.67 (95% CI, 0.48-0.92), respectively. None of the hospital factors measured was significantly associated with readmission. The Wald test revealed that GTP played a larger role than GTV in predicting hospital mortality (p = 0.018 vs p = 0.048) and FTR (p = 0.015 vs p = 0.985). Conclusions: Treatment at hospitals with higher GTP is associated with lower hospital mortality and FTR among geriatric patients. These findings suggest that creation of specialized services for geriatric trauma care can improve outcomes among geriatric trauma patients.