Racial/Ethnic Differences in the Prevalence of Clinically Recognized Alcohol Use Disorders Among Patients from the U.S. Veterans Health Administration

Background: Alcohol use disorders (AUDs) are common and have worse consequences for racial/ethnic minority groups than whites. AUDs are often underrecognized in clinical settings, but it is unknown whether the prevalence of clinically recognized AUD varies across racial/ethnic groups. We describe the overall and age- and gender-stratified prevalence of clinically documented AUD across 3 racial/ethnic groups in a national sample of Veterans Health Administration (VA) patients. Methods: Data from VA's National Patient Care Database identified all patients who used VA care in Fiscal Year 2012 and were documented as black, Hispanic, or white race/ethnicity. The prevalence of clinically recognized AUD based on ICD-9 diagnoses was compared across racial/ethnic groups overall and within gender and age groups using chi-square tests of independence. Results: Among 4,666,403 eligible patients, 810,902 (17.4%) were black, 302,331 (6.5%) were Hispanic, and 3,553,170 (76.1%) were white. The prevalence of clinically recognized AUD was 6.5% overall, and 9.8% (95% CI 9.8 to 9.9) among black, 7.1% (95% CI 7.0 to 7.2) among Hispanic, and 5.7% (95% CI 5.6 to 5.7) among white patients (p < 0.001). This pattern generally held for men, regardless of age group, with the exception of those 18 to 29 years old, for whom no difference was observed across race/ethnicity. Among women, the prevalence of AUD was generally lowest among Hispanic and highest among black patients, with the exception of those 30 to 44 years old, for whom the highest prevalence was among whites. Conclusions: In contrast to findings from the general population, the prevalence of clinically recognized AUD among VA patients is generally highest among black men and women and lowest among white men and Hispanic women. This is the first study to describe the prevalence of clinically recognized AUD across racial/ethnic groups in a large healthcare system. Future research comparing estimates to diagnoses based on structured gold-standard assessments is needed to understand whether AUDs are under- or overidentified.