Breast cancer is the second leading cause of cancer related death in the United States. Early detection through screening mammography is critical to reduce mortality. Previous studies have found disparities in the rates of mammography use between minority ethnic groups and non-Hispanic Whites. This research addresses the inadequacy of academic research on methods to predict screening mammography utilization among Asian American women (AAW). The cause for concern is that breast cancer incidence rates for AAW are increasing, while the rates are either stable or decreasing among other ethnic groups. The purpose of this study is to provide a means to reduce breast-cancer-related mortality rates among AAW. The theoretical framework in this study is the predisposing, reinforcing, and enabling constructs in educational diagnosis and evaluation model that groups the use of health services as a function of predisposing, reinforcing, and enabling factors. The research questions focused on identifying inputs, topological parameters, and techniques to build an optimal neural network prediction model. This quantitative study used California Health Interview Survey data of AAW, aged 40 years and above (N = 1850), which consists of variables such as health insurance and physician recommendation. Logistic regression was used to identify the predictors of adherence to mammography within 2 years. Results of this study showed that there are 11 inputs to an optimal prediction model, including physician recommendation, physical activity, and insurance status. A model based decision support system whose predictive accuracy for nonadherence was 88.46%, provided a framework that can identify populations at high risk of nonadherence. Implications for social change include providing intervention programs for early cancer detection via mammography to reduce mortality rates
