A population-based survival assessment of categorizing level III and iv rural hospitals as trauma centers

Context: Patients injured in rural areas are hypothesized to have improved outcomes if statewide trauma systems categorize rural hospitals as Level III and IV trauma centers, though evidence to support this belief is sparse. Purpose: To determine if there is improved survival among injured patients hospitalized in states that categorize rural hospitals as trauma centers. Methods: We analyzed a retrospective cohort of injured patients included in the Healthcare Cost and Utilization Project's Nationwide Inpatient Sample from 1997 to 1999. We used generalized estimating equations to compare survival among injured patients hospitalized in states that categorize rural hospitals as Level III and IV trauma centers versus those that do not. Multivariable models adjusted for important confounders, including patient demographics, co-morbid conditions, injury severity, and hospital-level factors. Findings: There were 257,044 admitted patients from 7 states with a primary injury diagnosis, of whom 64,190 (25%) had a "serious" index injury, 32,763 (13%) were seriously injured (by ICD-9 codes), and 12,435 (5%) were very seriously injured (by ICD-9 codes). There was no survival benefit associated with rural hospital categorization among all patients with a primary injury diagnosis or for those with specific index injuries. However, seriously injured patients (by ICD-9 codes) had improved survival when hospitalized in a categorizing state (OR for mortality 0.72, 95% confidence interval [CI] 0.53-0.97; OR for very seriously injured 0.68, 95% CI 0.52-0.90). Conclusions: There was no survival benefit to categorizing rural hospitals among a broad, heterogeneous group of hospitalized patients with a primary injury diagnosis; however the most seriously injured patients did have increased survival in such states.