It All Depends On How You Ask the Question
UR Measures Racial Disparity in Mammography Rates
Thursday, August 26, 2004
Black women may not be getting mammograms as often as white women, despite a widespread belief that this particular racial disparity has been overcome, a University of Rochester Medical Center study shows.
Study author Kevin Fiscella, M.D., M.P.H., is concerned that some doctors might be relying on a cursory view of national data that suggests the racial gap is closed. His analysis raises questions about the accuracy of mammography rates, which are important as medical authorities try to determine why the incidence of breast cancer is lower in black women than in white women, yet the death rate is higher among blacks.
Fiscella examined the methodology used by a large national survey of yearly mammography rates among 3,090 women ages 40 and older. He found that depending upon the way a survey question is phrased, the responses among black and white women can vary as much as 13 percent.
“How you ask a question clearly matters,” said Fiscella, an associate professor in the UR Departments of Family Medicine and Community and Preventive Medicine. The study is published in this month’s Preventive Medicine journal.
Using the 1996 Medical Expenditure Panel Survey, which is designed to assess national patterns of health care use and costs, Fiscella focused on two key questions.
The first question was, “How long has it been since you’ve had a mammogram?” Most women of all races and ethnicities tend to underestimate how long it’s been: They’ll circle 1 year, for example, when in fact it’s been 18 months or 2 years since their last mammogram. Indeed, the responses showed no significant differences in mammogram rates among black, white and Hispanic women.
But the responses changed to the question, “What health care services have you received in the last four months?” Fiscella found that when he looked at an entire year, 28 percent of white respondents versus 15 percent of black respondents reported receiving a mammogram. He believes the 13-percent disparity gives a more accurate picture of mammography rates.
Other studies have explained the tendencies of people while responding to surveys, Fiscella said. In some cases, minorities feel pressure to give a socially desirable answer, or feel threatened by preconceived notions about their health status. The second question is more neutral, because it simply asks for any medical services received.
The lesson, Fiscella said, is that the medical community should not rely exclusively on data reported by women themselves in order to monitor progress against disparities, but should confirm findings using other means.
# # #