BACKGROUND/RATIONALE:
A major challenge facing health services researchers is how to implement alcohol screening and brief intervention in real-world settings. For other preventive care primary care clinicians, quality managers and researchers use “markers” such as blood pressure and lipids, to screen for, assess the severity of, and track changes in risk over time. Although no valid lab marker of alcohol misuse is widely available, preliminary research suggests that the AUDIT-C, a brief validated alcohol-screening questionnaire, may be an excellent marker of alcohol-related risk.
OBJECTIVE(S):
The overall objective of the proposed study is to test whether the AUDIT-C is a marker for subsequent adverse health outcomes in a VA primary care population. We specifically evaluate whether the AUDIT-C predicts:
1) Medical conditions associated with excessive alcohol use (e.g. cirrhosis) over 5-7 year follow-up, and
2) Outcomes of three common chronic medical conditions, including process, short-term, and longer-term outcomes.
METHODS:
The study uses an existing VA data repository, which includes survey data; VA clinical, pharmacy, and laboratory data; VA and Medicare data for diagnoses; health care utilization; and vital status for 35,727 VA primary care patients who completed the AUDIT-C between 1997 and 2000. For Aim 1, we will test whether the AUDIT-C is a marker for subsequent development of 5 types of alcohol-related medical conditions using Cox proportional hazards regression. For Aim 2, we will conduct separate analyses of the AUDIT-C as a marker for adverse outcomes among patients with cardiovascular disease, chronic bronchitis and emphysema (COPD), and diabetes. These analyses will include 3 types of outcomes. We will use logistic or linear regression to test the association of the AUDIT-C with process measures (health behaviors and medication adherence). We will use linear regression to test the AUDIT-C’s association with two short-term outcomes (physiologic or laboratory measures and health status) and Cox proportional hazards regression to test the AUDIT-C as a marker of long-term outcomes (hospitalizations and mortality).
FINDINGS/RESULTS:
Analyses evaluating the association between alcohol screening scores and subsequent hospitalization for alcohol-related gastrointestinal (GI) conditions (bleeding and pancreatitis) were completed. Among 31,311 patients followed a median of 3.75 years, patients with AUDIT-C scores > 6 points were at significantly increased risk for GI hospitalizations. Secondary analyses demonstrated that the association was strongest among patients less than 50 years of age who reported drinking in the past year. These results suggest that scores on a brief alcohol screening questionnaire predict subsequent hospitalizations for alcohol-related GI conditions.
IMPACT:
The VA implemented alcohol screening with the AUDIT-C nationwide in 2004, and an electronic clinical reminder for AUDIT-C screening has been used over 1.5 million times since then. Soon AUDIT-C scores will be available in a national VA dataset. The proposed research is “calibrating” the AUDIT-C score as a measure of adverse health outcomes so that it can be used by VA clinicians, quality managers, and researchers as a marker of subsequent alcohol-related risks.
PUBLICATIONS:
Journal Articles
- Beste LA, Sun H, Maynard C, Bishop MJ, Henderson W, Schifftner T, Hughes G, Todd-Stenberg J, Kivlahan DR, Hawkins E, Harris A, DeBeneditti A, Bradley K. Prevalence of alcohol misuse in patients undergoing common surgeries in VA (Abstract). Journal of General Internal Medicine. 2008; 23: 377.
