A new national survey suggests that physicians’ experiences with patients, friends and family shape their breast cancer screening practices.
Share Fast Facts
Results of a national survey of more than 800 physicians suggest that their experiences with patients, family members and friends with breast cancer are linked with their recommendations for routine mammograms. Specifically, physicians who reported knowing at least one patient, family member or friend with a poor breast cancer prognosis and who had not been screened were more likely to recommend routine screening for their younger and older patients, age groups where routine screening is controversial.
A report of the findings, published Dec. 4 in JAMA Internal Medicine, highlights the impact physicians’ social networks may have on their adherence to nationally recognized breast cancer screening guidelines.
“Our findings suggest that we need to help clinicians better understand the impact personal experiences with friends and family members, as well as their patients, have on their practices,” says Craig Evan Pollack, M.D., M.H.S., associate professor of medicine at the Johns Hopkins University School of Medicine and the study’s lead author.
Current guidelines for younger and older women are discordant. For example, the American Cancer Society recommends personalized decisions for women ages 40-44, annual screening for women starting at age 45, and biennial screening for women 55 and older, whereas the U.S. Preventive Services Task Force recommends personalized decisions for women ages 40-49 and biennial mammograms for women 50-74.
For the new research, part of the Breast Cancer Social Networks study, Pollack and colleagues mailed surveys to 2,000 primary care physicians between May and September 2016. Participants were randomly drawn from the American Medical Association Physician Masterfile, a database of more than 1.4 million physicians, residents and medical students in the United States.
The survey asked respondents to detail experiences of two women they knew (one patient and one friend or family member) diagnosed with breast cancer, and to indicate whose cancer had the greatest impact on them. Each experience was categorized as: a) diagnosed through screening and with a good prognosis; b) not diagnosed through screening and with a good prognosis; c) diagnosed through screening and with a poor prognosis; d) not diagnosed through screening and with a poor prognosis; or e) unknown screening or prognosis.
Poor prognosis was defined as metastatic disease (disease that had spread) at the time of diagnosis or death from the cancer.
Respondents were also asked to report on whether they generally recommended routine screening mammograms to average-risk patients with no family history or prior breast issues in groups aged 40-44, 45-49 and older than 75 years.
A total of 848 physicians reported on 1,631 women they knew who had been diagnosed with breast cancer, 771 of whom were patients, 381 family members and 474 other social network members.
Of the physicians, 246 practiced internal medicine, 379 practiced family medicine and 223 practiced gynecology. The majority were males (461 of 848, or 54.4 percent) and non-Hispanic white (605 of 848, or 71.3 percent).
Pollack and colleagues found that physicians who reported at least one social network member with a poor prognosis without screening were significantly more likely to recommend routine screening to women ages 40-44 years (92.7 versus 85.6 percent) and older than 75 years compared to those who didn’t (84 versus 68.3 percent).
“As a first step toward increasing adherence to guidelines, it may be necessary to create opportunities and messaging strategies that help physicians recognize the experiences that help shape their recommendations,” says Pollack.
Other authors on this study include Kala Visvanathan of the Johns Hopkins Bloomberg School of Public Health, Archana Radhakrishnan of the University of Michigan, Andrew M. Parker and Sarah A. Nowak of the RAND Corporation, and Xinwei Chen of the University of Pennsylvania.
This study was funded by the National Cancer Institute (R21CA194194-02).
COI: At the time of the study, Pollack’s salary was supported by a grant from the National Cancer Institute.