Three year outcomes for American Indians/Alaska Natives in the Parent-Child Assistance Program in Washington State

Thesis (Master's)--University of Washington, 2015Objectives: The purpose of this study was two-fold: (1) to evaluate the associations between American Indian/Alaska Native (AI/AN) race/ethnicity and risk of a subsequent delivery of an infant prenatally exposed to alcohol/drugs at the conclusion of a three-year evidence-based prospective intervention designed to prevent future alcohol/drug exposed births among high-risk mothers who have substance abuse problems; and (2) to develop models predicting factors associated with increased risk among AI/AN women. Methods: This study included 795 women who were enrolled in the Parent-Child Assistance Program (PCAP) from 1999 to 2010 at nine PCAP sites in Washington State. Measurements were obtained from pre- and post-test interviews using the Addiction Severity Index and biannual assessment of substance abuse treatment received and alcohol and drug use. Statistical analyses conducted to compare AI/AN women to other racial/ethnic groups included bivariate analyses and logistic regression of various program and outcome measures. To determine which variables would predict lack of success in three PCAP outcome measures for AI/AN women, a backward stepwise logistic regression of selected factors was carried out. Results: Analyses compared 147 AI/AN women to 465 White, 101 Black and 62 women of other races. Pre-intervention factors varied somewhat among all racial/ethnic groups (AI/AN, White, Black, and the other races group). Bivariate analyses showed AI/AN women participating in PCAP to be younger than White and Blacks and to have more children and lower education compared to White women. Compared to women of the other racial/ethnic groups, a higher proportion of AI/AN women reported being separated from parents during childhood for two years or more, parental alcohol/drug abuse, and maternal alcohol use during pregnancy with them. Hierarchical logistic regression models comparing Whites to the other racial/ethnic groups indicated that AI/AN women did not have increased odds of “risk of subsequent exposed birth”. AI/AN women had increased odds of “not being abstinent from alcohol/drugs for 6 months or more at exit” versus Whites in the unadjusted model and in the model adjusting for demographics. This association was attenuated when additional confounders were added to the models. A comparison of AI/AN women to White women for the outcome “no regular family planning” showed no elevated odds of increased risk in adjusted models. In the stepwise models developed to predict risk of subsequent exposed birth, higher monthly income at exit and a high ratio of service needs met were protective for AI/AN women. Only age predicted not being abstinent from alcohol/drugs > 6 months, while prior inpatient treatment and having service needs met were protective for lack of regular family planning. Conclusion: In this study AI/AN race was not associated with an increased risk for having a subsequent alcohol/drug exposed birth compared to White women when other confounding factors were considered. Results from models focusing on risk factors for AI/AN women, however, suggest that meeting the service needs of AI/AN women may help prevent subsequent alcohol/drug exposed births among these participants