Regional variation in hospital mortality and 30-day mortality for injured medicare patients

Background We sought to evaluate how survival of older patients with injuries differs by geographic region within the United States. Methods We analyzed Medicare fee-for-service records for patients aged 65 years and older with principal injury diagnoses (ICD-9 800-959, excluding 905, 930-939, 958). Cases were classified by Maximum Abbreviated Injury Score (AISmax) and Charlson Comorbidity score (0, 1, 2, >= 3). Hospital mortality and 30-day mortality were modeled as functions of age, sex, AISmax, comorbidity, and geographic region (northeast, midwest, south, west). Results Hospital and 30-day mortality were both higher with male sex and increased age, AISmax, or Charlson score. Adjusted hospital mortality was highest in the northeast and south, but 30-day adjusted mortality was lowest in the same two regions. Conclusions Regional differences in risk-adjusted hospital survival for older patients with injuries are different from regional differences in 30-day survival. Hospital mortality as an outcome for older injured patients should be interpreted cautiously.