Predicting mammography utilization through model generation

Abstract

Currently, breast cancer screening with periodic mammography offers our best hope of reducing breast cancer mortality. The purpose of this report is to determine the best predictors of mammography screening using structural equation modelling with LISREL analysis. Using the Health Belief Model as a theoretical framework, the variables of perceived susceptibility, seriousness, benefits, barriers, health motivation, control, and suggestion by a health care provider, were hypothesized to directly impact mammography compliance with demographic/experiential variables indirectly influencing compliance. A total of 581 women selected through random digit dialling were interviewed in their homes by a graduate research assistant. A total of 51% of the compliance was predicted by direct and indirect paths. Variables that significantly predicted compliance with a direct path included health motivation, age, suggestion by a health care provider, knowledge and symptoms.