Baseline predictors of initiation vs. maintenance of regular mammography use among rural women

Abstract

Background

Predictors of regular mammography screening over many years have not often been examined prospectively. We used data from baseline (1993–1994), first (1996–1997), and second follow-up (2000) interviews with 336 White and 314 African-American rural women in the North Carolina Breast Cancer Screening Program to evaluate baseline factors predictive of regular mammography use over 7 years.

Methods

We defined regular mammography use as a recent mammogram (past 2 years) at all three interviews. Using binomial and logistic regression models adjusted for age, we examined factors associated with initiation (for women without prior regular use) and maintenance (for women with prior regular use) of mammography.

Results

Younger age and White race were predictive of initiation of regular mammography use. Physician recommendation was the strongest predictor of both initiation and maintenance of regular mammography use. Positive mammography attitudes and fewer personal barriers were strongly associated with initiation but not with maintenance.

Conclusions

Increased contact with providers and greater support for screening mammograms by providers could have an important impact on rural women initiating and maintaining regular mammography screening. Special efforts are needed to prompt rural African-American women and those over age 65 to initiate screening, since once they start they are likely to continue.

Introduction

Since the 1980s, the National Cancer Institute (NCI), the American Cancer Society (ACS), and other groups have recommended regular mammography screening at least once every 2 years for women between 50 and 74 years of age [1], [2]. Mammography use has increased dramatically since the mid-1980s. In 2000, 70% of women age 40 and older reported having a mammogram within the past 2 years, according to the National Health Interview Survey [3]. Although recent use of mammography screening has increased, regular screening, necessary to maximize breast cancer survival, remains low. The reported prevalence of regular screening (generally defined as obtaining two successive mammograms 1–2 years apart) has ranged from 28% to 57% between 1995 and 2001 [4]. Researchers have used a variety of measures of regular screening [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], making it difficult to obtain good estimates of rates of “regular mammography use”.

Most research designs used to examine mammography screening are problematic with respect to analysis of regular mammography use. Many studies employ cross-sectional designs to retrospectively examine factors at interview associated with mammography use [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17]. Cross-sectional designs prohibit disentangling initiation of mammography use from mammography screening maintenance; nor do they allow us to discern the extent to which attitudinal variables at interview either influence or are influenced by previous mammography use. Alternatively, retrospective cohort studies using HMO or other medical record data often lack detailed information on attitudinal or other psychosocial predictors [18], [19], [20]. To date, prospective designs have been limited to examining factors associated with recent rescreening over short intervals (1–2 years), often in the context of an intervention [21], [22], [23], [24], [25], [26], [27], [28]. A prospective cohort design such as that of NC-BCSP is necessary to evaluate, apart from the effect of an intervention, attitudinal and other psychosocial variables at baseline that may be predictive of long-term regular screening. Such a design also allows us to evaluate associations separately for initiation and maintenance of mammography use.

According to the Transtheoretical Model, interventions aimed at getting individuals to initiate a behavior should target removing barriers related to knowledge, attitudes, and beliefs, while interventions aimed at getting individuals to maintain a behavior should focus on supporting and prompting repetition of that behavior [30], [31]. Thus, a comparison of mammography initiation predictors with maintenance predictors may provide specific information on how best to target interventions that encourage women to obtain their first mammogram vs. those interventions designed to reinforce behavior already occurring.

Finally, while many studies have focused on urban women, to date none have examined regular mammography use among rural women; this population may face more access barriers to regular preventive care than do urban women.

We examined mammography use over a 7-year period in a cohort of older rural women from the control group of the North Carolina Breast Cancer Screening Program, as well as factors predictive of initiation and maintenance of regular mammography use by these women. The broad goal of these analyses was to provide information for more effectively developing and targeting future interventions for improving regular mammography use among rural women.