Unhealthy alcohol use (UAU) is common among HIV+ patients and associated with poorer adherence to antiretroviral therapy (ART), increased HIV disease progression, and poorer survival. Alcohol screening followed by routine brief intervention (BI) for screen-positive patients is widely recommended and now covered as a standard benefit under the Affordable Care Act. However, little is known regarding whether UAU is appropriately addressed in HIV+ patients receiving care, and there is currently no standard approach to monitor drinking changes over time in HIV+ patients. The Veterans Affairs (VA) Healthcare System is the largest provider of HIV care in the U.S. and is leading other systems in implementing alcohol screening and BI.
SPECIFIC AIMS. This study will make use of unique secondary VA data to: 1) Describe and compare receipt of BI among outpatients with UAU across HIV status; 2) Evaluate whether documented BI is associated with resolution of UAU among HIV+ outpatients with repeat alcohol screening; and 3) Evaluate whether changes in alcohol screening scores are associated with concomitant changes in HIV-related outcomes (ART adherence, markers of disease progression, and mortality risk based on the Veterans Aging Cohort Study (VACS) Index). APPROACH. This study applies validated algorithms and coding from VACS to unique VA data that includes repeated measures of alcohol screening and brief intervention in a cohort of ~ 7.4 million outpatients in care (1/2009 -12/2013).
Aim 1 analyses will be conducted among all patients who screen positive for UAU and uses logistic regression to estimate the prevalence and 95% confidence intervals (CIs) of documented BI for HIV+ and HIV- outpatients.
Aim 2 analyses will be conducted in the sample of HIV+ patients who screen positive for UAU and have repeat screening 12-24 months later and uses logistic regression to estimate the prevalence and 95% CIs of resolution of UAU for patients with and without documented BI.
Aim 3 analyses will be conducted among HIV+ patients who have initial and repeat screening 12-24 months later. Fixed-effect regression models will evaluate the association between changes in alcohol screen scores and changes in all scaled HIV-related outcomes.
Aims 1 and 2 analyses will be adjusted for patient-level covariates.
Aim 3 uses first-difference techniques, which implicitly adjust for all time-invariant subject-specific characteristics, including those that are measured and those that are not. IMPACT. This research uses unique national data with repeated measures of alcohol screening and BI to understand whether population-based implementation of alcohol screening and BI is reaching and effective for HIV+ patients and to determine whether routine clinical alcohol screening-which is increasingly being implemented into healthcare systems-can be used to monitor drinking outcomes in HIV+ patients for whom UAU is particularly risky. This research will inform other healthcare systems implementing alcohol screening and BI in response to healthcare reform and could lay a foundation for future comparative effectiveness research both within and across health systems.
The proposed innovative implementation evaluation uses a unique national dataset to understand whether population-based implementation of alcohol screening and brief intervention is reaching and effective for HIV+ patients, or whether more will be needed for this vulnerable population for whom unhealthy alcohol use is particularly risky. In addition, this study will evaluate whether routine clinical alcohol screening can be used to monitor drinking outcomes in HIV+ patients. This research will inform other healthcare systems implementing alcohol screening and brief intervention in response to healthcare reform, and may help lay the foundation for future comparative effectiveness research both within and across health systems.
