Factors Influencing Mammography Use Among Women in Medicare Managed Care

Health Care Financing Review, Summer, 2001 by Judith K. Barr, Susan Reisine, Yun Wang, Eric F. Holmboe, Karin L. Cohen, Thomas J. Van Hoof, Thomas P. Meehan

Three scores assessing Facilitators, Barriers, and Fear of Breast Cancer were calculated. As shown in Table 1, the participants scored fairly high on facilitating perceptions (mean = 62.5; SD=19.0), low on barriers (mean = 43.5; SD=18.1), and midrange on being fearful of breast cancer (mean = 49.9; SD=17.4). Relatively few women reported that their physician had never recommended a mammogram (mean = 34.4; SD=19.1), indicating that most women had received a physician recommendation at some point in their lifetime.

Seventy-two percent of the women had been with the plan for over 2 years, and a majority, 60 percent, had more than two visits to the physician in the past year. Most members (88 percent) were satisfied with their physicians (38 percent extremely satisfied; 50 percent satisfied). Three-quarters (76 percent) of the women were satisfied (28 percent extremely satisfied; 48 percent satisfied) with the health plan. The mean Access Problems Score was relatively low, with an average of 30.4 (SD=19.8), indicating that these women perceived relatively few problems accessing preventive health services.

Bivariate Analysis

Table 1 also presents the bivariate analysis of demographic characteristics, health status, health belief factors, health plan variables, and access with having or not having a mammogram in the past 2 years.

Age was significantly related to having had a mammogram in the past 2 years: the proportion not having a mammogram was 17 percent in the age group 65-69, 24 percent in the age group 70-74, and 59 percent in the age group 75 or over. Only 65 percent of the oldest women (75 or over) had a mammogram, compared with 87 percent and 86 percent in the younger age groups. Education also was significantly related to having a mammogram: those with an education of eighth grade or less education were less likely to obtain a mammogram. Health status as measured by the SF-12 was unrelated to mammography use, although those without a mammogram had a slightly lower mean score on the PCS and MCS than those with a mammogram. Of the health beliefs, only the Facilitators Score was significantly associated with having a mammogram, with those scoring high more likely to have had a mammogram. Having more visits to the physician, being satisfied with the physician, and perceiving fewer access problems to health care were significantly related to having a mammogram, as well.

Multivariate Model

Table 2 presents the results of the hierarchical multivariate logistic regression analysis. At the first step of the analysis, both age and education were significant factors. Those under age 75 were more than three times as likely to obtain a mammogram as older women. Those with some college education were more likely than those with less than a high school education to obtain a mammogram. At the second step with demographic and health status variables in the analysis, the effects of age and education remained largely unchanged. Health status was not significant, nor were there significant interaction effects between age and health status on mammography utilization. At the final step, taking into account all other independent variables, age remained significantly related to mammography use in this sample. Women in the age group 75 years or over were less likely to have received a mammogram in the last 2 years compared with younger women. Health beliefs also were important. Those who scored high on the facilitator measure were significantly more likely to have obtained a mammogram (OR=1.09; CI = 1.06-1.11; p [is less than] 0.001). The Cue to Action measure emerged as an important variable, as those who had never been told by a physician to obtain a mammogram were significantly less likely to report this service. Finally, satisfaction with physician remained important; those who were satisfied were more than twice as likely to have received a mammogram as those who were not satisfied (OR=2.68; CI=1.01-7.42; p [is less than] 0.05). Factors that were not significant predictors of mammography use in the multivariate model include education, health status, attitudinal barriers, fear of breast cancer, number of visits, and perceived access problems. The logistic regression model was a good fit, explaining 36 percent of the variance in reporting a mammogram in the past 2 years.