Health care plans and providers are increasingly aware that they must identify, treat, and manage people who have alcohol problems and comorbid conditions such as depression and chronic illnesses. However, to identify such patients and address their problems appropriately, we need to understand how these comorbid conditions interact, and how patient characteristics affect such conditions and their outcomes. Researchers have been working to understand the factors that influence alcohol consumption, and to learn how that consumption affects morbidity, mortality, and use of health services. To date, our understanding of these relationships remains limited because the processes that affect alcohol use and utilization of health services are complex. To advance our knowledge, models must now include a complex mix of mental and physical comorbidities, functional status and individual characteristics. One of the most important of the individual-level factors affecting alcohol consumption and health service use is gender: It affects patterns of alcohol consumption, amounts and types of medical care used, and comorbid psychiatric and health conditions that underlie both alcohol and service use. In this exploratory/developmental grant, we will use existing data sets to address four main questions: (1) How do demographic factors, depression, health and functional status, and chronic illness jointly influence alcohol consumptions (2) How do demographic factors, alcohol consumption, depression, health and functional status, and chronic illness influence total cost of health services? (3) Do gender-specific models corresponding to Questions 1 and 2 provide superior explanations when compared to combined-gender models? (4) Do structural equation models enhance our understanding of these complex phenomena when compared to two-part regression models? Using an existing data set of 7844 randomly sampled HMO members age 25 and older, we will construct comprehensive regression and structural equation models of the processes leading to alcohol consumption, and then use the same modeling strategies to study how alcohol consumption affects health services use and costs. We will compare gender-specific models with combined-gender models to determine if the underlying processes differ between men and women, and evaluate the knowledge gained from the two-part regression and structural equation models to assess their relative contribution to our overall understanding of these outcomes.
| Green, Carla A; Polen, Michael R; Perrin, Nancy A et al. (2004) Gender-based structural models of health care costs: alcohol use, physical health, mental health, and functioning. J Ment Health Policy Econ 7:107-25 |
| Green, Carla A; Polen, Michael R; Perrin, Nancy A (2003) Structural models of gender, alcohol consumption, and health. Subst Use Misuse 38:97-125 |
