“…Through a literature search, we identified several patient characteristics as a priori potential confounders and included those available from SHEP data as covariates based on their associations with receipt of brief interventions and patient-reported quality-of-care ratings. 31,34,37,[39][40][41] These included age, gender, race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, or other), educational attainment (less than high school, high school, General Educational Development Test [GED], or 2-year degree; 4-year degree; or greater than 4-year degree), and Short Form-1 self-reported health status (poor, fair, good, very good, or excellent). In addition, we included severity of unhealthy alcohol use, as previous studies have demonstrated associations between greater alcohol consumption and receipt of brief interventions, 37,39,40 and severity may also be associated with patient-reported quality-of-care ratings.…”