Bridging Disciplinary Boundaries (January 11 - 14, 2007)

Purpose: This study examined (1) low-income pregnant and/or parenting women's individual and social/environmental risk and protective factors and treatment program characteristics that influence their treatment retention and completion, and (2) identified interaction effects between client factors and program characteristics that may influence women's treatment retention and completion. The women's race/ethnicity (African American, Hispanic, and White) was examined as a major variable, as was the type of residential program in which they were treated. The residential treatment programs were of two types: those that allowed the women's children to accompany them into treatment and those that did not. Research indicates that pregnant African American women are at greater risk for alcohol consumption and have poorer treatment retention and completion rates (Ashley et al., 2004; Mertens & Weisner, 2000; Noble et al., 1997). However, there is a paucity of research about pregnant and parenting women of color, especially Hispanics. Methods: Hierarchical linear modeling was used to analyze client-level (individual and social/environmental) data and program-level data. The client data (n=13,117) were secondary data obtained from the Texas Department of State Health Services, Substance Abuse Services, and are subject to quality control procedures. The sample included 25% African American, 29% Hispanic, and 44% White women. The program data were primary data collected by the researcher from 28 program sites [14 allowing children to accompany their mothers into treatment (which served 10,423 women in the study) and 14 that did not (which served 2,694 women in the study)] over a five-year period (June 1, 2000, through May 31, 2005). Examples of the 41-item program questionnaire are Number of beds for mothers and children and types of services provided. To control for client demographics, service delivery, and resource availability, the 14 programs that do not allow children to accompany their mothers were within a 10-mile radius of the 14 programs in the state of Texas that allow children to accompany their mothers. These are the largest client samples ever used to examine a population of mothers in chemical dependency treatment. Results: Interaction effects revealed that unlike their White counterparts, minority women were less likely to complete treatment if they were pregnant (African American), referred by family or friends (Hispanic), exposed to drugs at home, older when they began using their primary substance, treated by master's level clinicians, or visited the emergency room/hospital more. Protective factors for minority women included employment, higher incomes, older age, and if enrolled in the program with their children accompanying. White women were less likely to complete treatment if they had a co-occurring disorder, were homeless, or were in a facility with a larger number of women. All ethnic groups improved when they had more education and remained in treatment longer. Implications for practice and policy: This study identified individual, social/environmental, and program factors that may increase/decrease a woman of color's ability to complete treatment. This knowledge can improve providers' treatment strategies. Additionally, these results reinforce the need for additional funding to increase a client's duration in women's residential chemical dependency programs.