A Multisite Randomized Controlled Trial of VA Integrated and Enhanced Referral Behavioral Health Models on Alcohol Misuse in Older Male Veterans
METHODS: This study was a secondary analysis of data collected at five Department of Veteran Affairs (VA) Medical Centers as part of the Primary Care Research in Substance Abuse and Mental Health for Elderly (PRISM-E) study, a multisite randomized controlled trial. The sample included 1222 older male veterans who were randomly assigned to integrated or enhanced referral behavioral health care at VA primary care clinics with 6 to 12 month follow-up. Brief alcohol treatment was provided to at-risk drinkers by behavioral health providers either co-located within primary care clinics (integrated treatment model) or in a separate location offering transportation assistance and coordinated care between behavioral health providers and primary care physicians (enhanced referral model). Alcohol misuse included both at-risk drinking and problem drinking. At-risk drinking for older adults was defined as more than one drink daily, and problem drinking was assessed using the Short Michigan Alcoholism Screening Test-Geriatric Version (problem drinking = > 3 affirmative responses). Both behavioral health models required a minimum of one treatment session, with no maximum number of sessions over a 6-month period. After 6 months, attendance at treatment sessions was voluntary.
RESULTS: At baseline, 35.8% (n = 438) reported at-risk drinking and of those 46.1% (n = 200) reported problem drinking. At 6 months, 21.1% (n = 234) reported at-risk drinking and of those 41.4% (n = 150) reported problem drinking. At 12 months, only 36.3% (n = 89) reported problem drinking. Bivariate analyses revealed no significant differences between behavioral health models in at-risk and problem drinking over time. Generalized linear mixed modeling revealed treatment and time effects were significant for at-risk drinking after controlling for VA site, race, education, employment, and depressive symptoms. Treatment and time effects were significant for problem drinking after controlling for VA site, race, and depressive symptoms. The treatment x time effect was non-significant for at-risk and problem drinking.
CONCLUSIONS: Both integrated and enhanced referral behavioral health models were effective in reducing alcohol misuse among older male veterans seen in VA primary care clinics with follow up over 12 months. Because alcohol misuse complicates chronic health conditions and prescription drug reactions, screening and brief intervention for alcohol misuse in primary care can be cost-effective and potentially result in a medical cost offset. Based on primary care protocols, VA health policymakers should determine which behavioral health model is best for their patients, providers, and infrastructure. Future research should examine treatment processes of both behavioral models to determine which are most effective in reducing alcohol misuse in older veterans.