Predicting health care utilization in marginalized populations: Black, female, street-based sex workers

Abstract

Background: Patterns of social and structural factors experienced by vulnerable populations may negatively affect willingness and ability to seek out health care services, and ultimately, their health.

Methods: The outcome variable was utilization of health care services in the previous 12 months. Using Andersen's Behavioral Model for Vulnerable Populations, we examined self-reported data on utilization of health care services among a sample of 546 Black, street-based, female sex workers in Miami, Florida. To evaluate the impact of each domain of the model on predicting health care utilization, domains were included in the logistic regression analysis by blocks using the traditional variables first and then adding the vulnerable domain variables.

Findings: The most consistent variables predicting health care utilization were having a regular source of care and self-rated health. The model that included only enabling variables was the most efficient model in predicting health care utilization.

Conclusions: Any type of resource, link, or connection to or with an institution, or any consistent point of care, contributes significantly to health care utilization behaviors. A consistent and reliable source for health care may increase health care utilization and subsequently decrease health disparities among vulnerable and marginalized populations, as well as contribute to public health efforts that encourage preventive health.