*187. Racial Differences in Health Care Use within a Brief Intervention Study to Reduce Drinking
LA Copeland, Ann Arbor VAMC HSR&D Field Unit
Objectives: To estimate the effect of race/ethnicity (African American vs. Caucasian) on VA health care utilization among older male drinkers enrolled in a brief intervention to reduce drinking.
Methods: Veterans (N=192) aged 55+ who drank 12+ drinks/week or binged 1+ times in the past 3 months completed a baseline survey for a large, primary care intervention study to reduce drinking.Veterans were randomized into treatment and control conditions (Tx n=92; Ctl n=100). Surveys were linked to VA data. Types of care included inpatient days and outpatient clinic stops -- psychiatric, alcohol and other drug-related (AOD), medical, and total -- plus inpatient nursing home days and outpatient miscellaneous (screening, therapy) stops. Utilization was summed over both 9 and 18 months, pre- and post-intervention. Repeated measures MANCOVA controlling for age analyzed the effects of race/ethnicity on utilization before and after the survey/intervention.
Results: The sample comprised 46 African American and 146 Caucasian veterans. African American veterans differed from Caucasians: 3.6 years younger, more likely to live alone (48% vs 30%), scored higher (worse) on SMAST-G (3.6 vs 2.1) and CAGE (2.1 vs 0.9) scales assessing probable alcohol dependence. No differences were apparent on education, average drinks per week, or binge drinking.Pre-intervention health care usage was generally similar between the racial/ethnic groups, but differed over time for some types of care. In the nine months after baseline, AOD outpatient clinic stops increased more for African American veterans (from 0.04 to 1.20 stops) compared to Caucasians (0.01 to 0.02). In the 18-month follow-up, medical stops (Afr: 12.6 to 17.5 vs Cau: 12.5 to 13.4) and total stops (Afr: 25.5 to 39.6 vs Cau: 24.8 to 24.5) also showed this effect. In addition, miscellaneous outpatient stops and hospitalizations had significant three-way interactions of time*race/ethnicity*treatment, reflecting an analysis of four groups over time: Cau-Ctl, Cau-Tx, Afr-Ctl, and Afr-Tx veterans. Initial miscellaneous clinic stops ranged from 9.8 to 12.5 for the four groups, but in follow-up African Americans averaged 21.3 stops while other group means were 11.0-11.7 stops. Number of hospitalizations was lower for African Americans initially (Tx=0.05, Ctl=0.08; compare to Cauc. Tx=0.28, Ctl=0.19) but rose differentially such that, after receiving the intervention, African Americans appeared to be hospitalized most frequently (Afr: Tx=0.63, Ctl=0.21; compare to Cau: Tx=0.22, Ctl=0.33).
Conclusions: In older veterans who drank in excess of NIAAA limits, a brief outpatient intervention to reduce risky drinking was associated with greater increases in health care use by African American veterans. Increased use by African Americans may be a result of: (1) events within the initial (recruitment) outpatient visit, (2) factors not measured such as comorbidity or intensity of alcohol abuse, (3) social factors affecting when African Americans use health care services, or (4) a greater impact from participating in the study or receiving the intervention.
Impact: In this sample of at-risk drinkers, changes in health care use were associated with race/ethnicity and treatment. Within sub-populations, brief interventions could be used to identify or adjust health care disparities commonly seen between these two racial/ethnic groups.