2010. Effects of Ethnicity on One Year Mortality in Seriously Ill Veteran Population
SM Fischer, Denver VAMC and University of Colorado Health Sciences Center, AS Sauaia, University of Colorado Health Sciences Center, A Kramer, University of Colorado Health Sciences Center

Objectives: Despite national efforts to reduce health disparities, the inequalities in care and health outcomes for ethnic minorities continue to grow. Studies in the VA system have demonstrated that despite equal access and care, African Americans present with more advanced disease and have higher all cause mortality. o our knowledge, there have been no similar studies including Latinos. This study was designed to determine whether Latino and African American ethnicity is an independent predictor of one-year mortality following an index hospitalization.

Methods: Study Design: A retrospective cohort study to determine the independent effect of Latino and African American ethnicity (defined as phenotypic appearance as recorded in the medical record) on a logistic regression predictive model including the CARING criteria (Cancer, Admissions > = 2, Residence in a nursing home, ICU admit with MOF,  > = 2 Non-cancer hospice Guidelines).These criteria were developed and validated in a population of seriously ill hospitalized Veterans to predict one-year mortality. Mortality was also adjusted for age. Sex was not included in the model because 98% of patients were men. Setting: Denver Veterans’ Administration Medical Center (DVAMC). Patients: All patients admitted to the medical wards from February 1, 1999 through March 31, 1999 (n = 435).

Results: The study population included 9% Latinos and 13% African Americans, while Denver’s overall population is 20% Latino and 5% African American. Forty-seven percent of the hospitalized Veterans met at least one of the CARING criteria. Latino ethnicity and each CARING criteria were significant predictors of mortality in unadjusted chi-square tests. Latino ethnicity had an independent effect on one-year mortality when this variable was entered into a predictive model with an adjusted odds ratio of 2.5 (95% CI 1.1 – 5.7). Adjusted mortality was not higher among African Americans (adjusted odds ratio 0.8 (95% CI 0.3 – 1.9).

Conclusions: Latino but not African American ethnicity had an independent effect on one-year mortality for hospitalized seriously ill adults in an equal access system.

Impact: Latino but not African American ethnicity had an independent effect on one-year mortality for hospitalized seriously ill adults in an equal access system.