This study assessed whether health beliefs and concerns have independent effects on screening mammography utilization when race/ethnicity and other important patient-related predictor variables are controlled. Hypotheses were developed based on the Health Belief Model.

Data were collected in 1991 as part of a survey of Medicare beneficiaries; patients included were those from the California site who did not have breast cancer. The original study was an intervention trial regarding physician education, and both arms of the trial were included in this study.

The whites in this sample were well educated (almost 90% completed high school) and had adequate insurance. These characteristics were not true for the blacks (less than 60% completed high school and 85% did not have adequate insurance) nor for Hispanics (just over 30% completed high school and more than 90% did not have adequate insurance).

Beliefs about mammography were similar across racial groups, except a higher proportion of whites than the other two groups believed it detected cancer. Concerns varied across racial groups: blacks were most concerned (about all three measures, radiation, pain and cost), and whites were least concerned.

In multivariate models to predict mammogram within the last two years (which included sociodemographic characteristics, SES, regular care source, transportation, benign breast disease, health and functional status, knowledge and beliefs), the inclusion of a variable for racial group did not alter the influence of knowledge and beliefs. The belief that mammogram eases the mind and a history of benign breast disease, in addition to age, source of health care, and good health, were significant predictors of having a mammogram.

Although this study was not designed to test the effect of race per se, it did demonstrate that after controlling for various patient-level factors, African American and Hispanic women were not more or less likely to receive a mammogram than white women. The authors suggested this is in accordance with studies that show SES is a more important predictor of utilization in recent years. Thus, the absence of racial differences in this study might be explained if SES factors are less important in this elderly population than in other samples. The authors argued that SES is less important in this sample because, "although over half of the African Americans (54%) expressed concern about the cost of a mammogram, over 95% of the respondents across all ethnic groups were Medicare enrollees and 65% had supplemental insurance."