Abstract
Within the Bemidji Indian Health Service Area (Minnesota, Wisconsin, Michigan, Indiana) women's adherence to mammographic screening guidelines is low. This pilot study's purpose was to examine whether compliance with annual mammogram screening is associated with the presence of risk factors used in Gail Model calculations. Risk factors were consistently recorded on patients' self-history questionnaires. Data from 203 consecutive Native women's records were reviewed (≈ 17% of the ultimate sample of 1200 records). All records were from one tribe. Inclusion criteria were: ≥ age 40, no prior history of breast cancer, and ≥ one mammogram on file. Data reviewed included current age, breast cancer family history, personal biopsy history, ages at first live birth, menarche and menopause, and compliance with annual mammogram. Women averaging ≤ 1.2 years between mammograms were considered compliant. Descriptive statistics were calculated to categorize women as being at higher or lower risk for each breast cancer risk factor. Fisher's Exact two-tailed test measured associations between screening compliance-noncompliance and women's risk on each factor.
Approximately two-hundred records were used. Six were excluded due to missing data. Excepting the mean age (46.1) at menopause, mean ages overall, at menarche, and first live birth were consistent with the United States (US) general population. Menopause occurred approximately 5 years earlier in this sample. The median age at menopause in this group was 48 years. No statistically different relationships were found between screening compliance and: breast cancer family history (p=1.00, n=202); age at menarche (p=0.31, n=195); and age at menopause (p=0.39, n=198). There was a statistically significant relationship between screening compliance and age (p=0.012, n=187) based on whether women were at or above the median age; breast biopsy history (p=0.007, n=192); and age at first live birth (p=0.005, n=185). The latter association showed women who were younger at first live birth were more likely to adhere to screening guidelines than were women who were at higher risk due to later age at first live birth.
Preliminary results showed that having a higher risk factor was not predictive of screening compliance in half of the six measures, and in one case, (age at first live birth) the association was not in the direction anticipated. Absence of expected associations may be related to inadequate awareness of personal risk and screening.
A larger study should further delineate elements contributing to poor screening compliance.
This study highlights potential areas that providers may influence to increase screening compliance.
The study was supported with Native American Research Centers for Health funding through the Great Lakes Intertribal Council.
Second AACR International Conference on the Science of Cancer Health Disparities— Feb 3–6, 2009; Carefree, AZ