Mammography Use--Wisconsin, 1980-1993

In the United States, efforts to reduce mortality from breast cancer focus primarily on secondary prevention (ie, early detection and treatment). Since 1980, private, public, and voluntary organizations in Wisconsin have promoted screening mammography as a means for reducing the death rate from breast cancer. To assess the effectiveness of these efforts, the Division of Health, Wisconsin Department of Health and Social Services (DOH), analyzed data from annual statewide surveys of mammography providers during 1989 through 1993 and data about self-reported mammography use from the Behavioral Risk Factor Surveillance System (BRFSS) during 1987 through 1993. This report summarizes these analyses and trends in the number of mammograms performed annually in Wisconsin during 1980 through 1993.

An annual survey of all registered mammography providers in Wisconsin has been conducted since 1989. During 1980 through 1993, the number of mammography providers ranged from 76 to 236. Survey questionnaires are mailed annually to all mammography providers in conjunction with a mailing of DOH radiation-protection registration materials. The questionnaire asks each facility to estimate the total number of mammograms performed during that year and to provide information about mammography referral and follow-up procedures, fees, and availability of low-cost screening services.

The 1989 survey requested estimates of the total number of mammograms performed during 1980 through 1989. The response rates for the five surveys con ducted during 1989 through 1993 were 75%, 89%, 90%, 91%, and 91%, respectively. Data were adjusted for nonresponse to provide statewide estimates of the annual number of mammograms performed.

Trends in self-reported mammography use were determined by analyzing data from the BRFSS, which has included questions about mammography use since 1987. The BRFSS is a random-digit-dialed telephone survey of household residents age 18 years that provides population-based surveillance data about selected health behaviors. The total number of mammograms performed in the state each year during 1987 through 1993 was estimated from the BRFSS by multiplying the adult female population in Wisconsin by the proportion of women who reported having had a mammogram during the preceding 12 months. The number of adult women interviewed in Wisconsin for the annual BRFSS ranged from 673 (in 1990) to 857 (in 1993).

Analyses of the mammography provider surveys (1989 through 1993) and the BRFSS (1987 through 1993) indicated steady increases in the number of mammograms performed each year during 1980 through 1993. Based on the provider survey, the total number of mammograms performed each year increased nearly 17-fold, from 31,000 in 1980 to 517,000 in 1993. Compared with the provider survey, estimates based on the BRFSS were consistently higher, varying from 81% higher in 1988 to 17% higher in 1993. In addition, the number of mammography providers increased 310%, from 76 in 1980 to 236 in 1993. Standard errors for annual BRFSS estimates ranged from ±1.5% to ±1.7%.

Editorial Note from the CDC: The findings in this report document a nearly 17-fold increase in the annual number of mammograms performed in Wisconsin during 1980 through 1993. Although the estimated number of mammograms performed each year differed substantially by data source, the trends were similar for both sources. Two important factors probably contributed to the increase in mammography use during this period: (1) the substantial increase in the acceptance and use of screening mammography among physicians during the 1980s, and (2) initiation of extensive efforts to educate the public and health-care professionals about national screening mammography guidelines, which were implemented during the late 1980s. Other possible contributing factors include the initiation of low-cost mammography screening programs and wider availability of high-quality, low-cost mammography equipment beginning during the early 1980s.

The findings in this report are subject to limitations associated with the two data sources. Although the provider survey is an example of a low-cost, efficient method to characterize trends in mammography use, it does not collect information about sociodemographic variables and may overestimate mammography use because data were based on the number of mammograms performed and some women receive more than one mammogram each year. In addition, the estimates based on the self-reported BRFSS data probably were higher than those from the provider survey for at least three reasons. First, some women who reported having had a mammogram during the previous 12 months probably received the mammogram more than 12 months previously. Second, women who participated in the BRFSS may not be representative of the total population of women in Wisconsin because they had telephones, were better educated, or were more likely to have had insurance coverage-factors related to an increased likelihood of having received a mammogram. Third, the BRFSS is a household survey that does not include institutionalized women (eg, those in long-term care facilities), who are less likely than noninstitutionalized women to receive mammograms.

The findings in this report are being used in Wisconsin to further identify groups of women who underuse mammography screening, develop intervention strategies to increase mammography use, and assess progress toward the year 2,000 national health objectives for breast cancer and mammography (objectives 16.11 and 16.16). This approach can be adapted for use by health departments in other states to assess the effectiveness of efforts to promote mammography screening.

Adapted from Morbisity and Mortality Weekly Report 44(41):754-756, 1995.