News Release

Most older women don’t get mammograms often enough, but mailed reminder can boost use

Pair of new studies looks at general Medicare-aged population

Peer-Reviewed Publication

Michigan Medicine - University of Michigan

ANN ARBOR, MI – More than half of women over the age of 65 who should be getting regular mammograms aren't, a new study finds.

But a companion study shows that a simple mailing -- reminding these women of their Medicare coverage for the exam, and of the high risk of breast cancer they face due to their age -- is enough to prompt a meaningful increase in mammogram use.

The two studies, published simultaneously this week by researchers from the University of Michigan Health System and U-M School of Public Health, show that older women may need an extra push from their doctors, families and Medicare administrators to go get a screening test that may save their lives.

And, the researchers say, that extra push can be as simple as an inexpensive but personalized mailing made possible by Medicare's computer database systems.

"Older women seem to be using their age as a reason to stop getting mammograms, but we know that their risk of breast cancer – and the effectiveness of mammography in identifying it early – rises with age," says lead author R. Van Harrison, Ph.D., an associate professor in the U-M's Department of Medical Education. "We need interventions that can help spur women over 65 to get this needed screening. Personalized mailings and outreach to physicians should help."

The U-M team performed the two-part research using Michigan's computerized Medicare database and funding from the National Cancer Institute.

First, they looked back at all mammogram use from 1993 to 1997 among 10,000 randomly sampled Michigan women over the age of 65. Those results are published in the March issue of Cancer.

Then, they performed a controlled study to see if a carefully designed personalized mailing would boost mammogram rates among women who hadn't had the exam in the last five years. The data from that study are in the March issue of Medical Care.

The mammogram use study found that 43 percent of the women had not had a single exam in the five years of the study, and 17 percent had had only one. The team then sent a personalized letter to more than 1,200 women over 70 years of age who hadn't had a mammogram in five years, urging them to schedule one. Though the absolute increase in this group's mammogram rate was modest, they were 60 percent more likely to get an exam than 1,200 similar women who didn't get a letter.

A women's risk of breast cancer rises throughout her life, and women in their 60s, 70s and 80s have the highest risk of all. Fifty-nine percent of breast cancer deaths occur in women aged 65 or older. Meanwhile, mammograms are most accurate in women of these ages, with lower chances of anxiety-raising false positives because older women's breasts are less dense.

Mammography utilization study summary

The Medicare database analysis revealed several factors that were more common among the 67 percent of Michigan women who had at least one mammogram in the five years of the study. On the whole, the younger a women was, and the fewer times she had been admitted to the hospital for any cause, the more likely she was to have had at least one mammogram.

Women who had one or more mammograms in the five-year period were also much more likely to have more than one doctor involved in their care, and more likely to have seen gynecologists, who may be highly attuned to the importance of mammograms for women over 65 years of age. Some of the women with no mammograms in five years were found to have Medicare claims for medical or mental health conditions that may preclude their having a mammogram.

The study was specially designed to look for any impact that race, urban or rural location, education level and income might have on mammogram rates, but no impact was found. Says Harrison, "This may reflect a success story of public awareness and Medicare coverage."

Based on their findings from mammogram use study, the U-M team made recommendations for efforts to increase screening among older women.

Awareness campaigns, they write, should work to dispel the perception that older women don't need mammograms, and should emphasize the increased risk of breast cancer facing older women, and the value of mammograms for older women with a reasonable life expectancy. Specific interventions, they added, should target women whose Medicare files show no recent mammograms, and aim to help physicians do better at recommending mammograms.

Intervention study summary

Harrison and his colleagues then set out to test a method of targeting older women with no mammograms in the past five years, and no medical or mental health conditions likely to interfere with their ability or need to have a mammogram. From the Medicare database, they selected 1,226 pairs of these women, matching them by race and zip code of residence. The study oversampled rural and African American women.

One woman from each pair was sent a stamped, personally addressed Medicare envelope containing a letter on Medicare stationery from Medicare's Michigan medical director, a brochure designed by U-M health education experts emphasizing the importance and availability of mammography for older women, and a sheet detailing the steps to schedule a mammogram. (Faxed copies of these materials are available to media upon request.)

The letter was personalized, meaning that it was addressed to the woman by name ("Dear Jane Smith"), and mentioned her own mammogram use history ("Your insurance records show that you have not used this benefit recently"). The brochure showed women of different ethnicities, and contained testimonials on the importance of a women's good health to the rest of the family. All the materials were written at an 8th grade reading level, printed in large type, and assessed by focus groups of older women of different ethnicities, and by health experts.

The researchers sent the letters to the women in the intervention group, and sent nothing to the women in the control group. They then looked back at all the women's Medicare files 14 months after the letter was sent, and analyzed mammography use and the cost of the intervention.

Only five percent of the women who did not receive a letter had a mammogram in the follow-up period, compared with 8.1 percent of the women who received a letter. This difference, though only 2.9 percentage points, represents a 60 percent increase in mammography use.

As in the previous study, younger women were more likely to have a mammogram than older women, but most of the effect of the letter was seen in women in their 70s – a 4.1 percentage point increase from 6.5 percent to 10.6 percent. Most of the effect was in the first six months.

Notably, Harrison says, most of the increase in mammograms received by women who got the personalized letter was due to an increase in diagnostic mammograms. These more detailed exams, which are only paid by Medicare for women who have a palpable lump in their breast or a major non-age risk factor for breast cancer, are more thorough than screening mammograms.

"The finding that most of the increase was due to an increase in diagnostic mammograms was a surprise," says Harrison. "It shows that the intervention seems to work best on women who already had a reason to be nervous about their breast cancer risk. In other words, it's pushing the women who may be considering a mammogram to go ahead and get one."

This may make personalized intervention even more cost-effective than experts might expect, Harrison notes. "Those who do change appear to be those whom it's most important to change."

The mailing could be done on a statewide or nationwide basis relatively inexpensively, the authors say. For instance, more than 150,000 Michigan women age 70 or older in 1997 might have met the criteria to receive the mailing. It would have cost the state $176,459 in materials, postage and staff time to reach them. That would have resulted in 4,360 more mammograms.

In addition to Harrison, the authors of the studies include Laurence McMahon, Jr., M.D., M.P.H., chief of general medicine in the U-M Medical School's Department of Internal Medicine, who considered medical issues and helped oversee the large-scale data analysis; Nancy Janz, Ph.D., a School of Public Health associate professor who led the design of the intervention materials; Jeoffrey Stross, M.D., professor of General Medicine, who helped consider medical issues; Michael Chernew, Ph.D, School of Public Health and General Medicine associate professor, who designed the economic analysis; Robert Wolfe, Ph.D., professor, SPH, who oversaw the statistical design; Philip Tedeschi, Ph.D., professor, SPH, who oversaw analysis of the Medicare claims data; and Xuelin Huang, Ph.D., formerly SPH, now at the M.D. Anderson Cancer Center, who helped design and performed much of the analysis.

Background on mammography recommendations for older women:

Despite the recent controversy over the benefits and risks of mammography for women in their 40s, there has been little debate over the value of regular mammograms for women over 50.

Since 1986, the federal government's U.S. Preventive Services Task Force has officially recommended that all women ages 50 to 70 have mammograms every year or two. One year ago this week, they extended that recommendation to women in their 40s, while emphasizing that available data show women in their 50s and 60s get the most life-saving benefit from the exam.

Even though the 2002 USPSTF panel had no specific data on mammography in women aged 70 and up, they projected results from studies on women under 70 years of age and recommended that women over 70 who had no life-threatening diseases get mammograms every year or two. NCI recommends the same.

Meanwhile, Medicare Part B has covered most of the cost of a screening mammogram every two years since 1991, and every year since 1999, for women aged 65 or older. Financial assistance is often available for low-income women who cannot pay the remainder of the cost.

Mammogram rates have risen in recent years, due to Medicare coverage and public awareness campaigns about the benefits of regular mammography. But Harrison and his colleagues note that their study's mammogram rate, based on actual Medicare claim data, is both lower and more reliable than results of studies based on women's self-reported screening histories.

###


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.