Patient Factors Associated with Breast Cancer Screening Among Older Women

OBJECTIVES: To identify patient factors that influence physicians' decisions to recommend screening mammograms for older women. DESIGN: A cross-sectional survey including clinical case vignettes. PARTICIPANTS: Random sample of Massachusetts internists, obstetrician/gynecologists, family/general practice physicians, and geriatricians. OUTCOME MEASURE: Proportion of screening mammograms recommended for women in different versions of each case vignette. RESULTS: A total of 482 (65%) of the eligible participants responded to a mailed survey of questions about breast cancer screening practices, attitudes toward ACS guidelines, and four clinical vignettes. Vignettes tested the impact of patient's age, cognitive function, nursing home residence, functional limitations, and comorbidity on the physician's decision to recommend a mammogram. Ninety-four percent of physicians reported often performing periodic clinical breast exams and mammograms for women aged 65 to 74. For women aged 75 to 84, 89% of physicians reported often performing periodic clinical breast exams, and 79% recommend mammograms. Only 48% strongly agreed with ACS guidelines for annual mammography for women over 65. Age, dementia, and nursing home residence were patient factors associated with decreased mammogram use, but limited mobility and chronic medical problems were not. Physicians were more likely to recommend a mammogram for a woman aged 65 to 74 than for a woman 75 to 84. (P = .002) Physicians were more likely to recommend a mammogram for a woman without dementia than for a woman with mild dementia (P < .05) and for a woman living with her daughter than a for a woman living in a nursing home (P < .001). CONCLUSIONS: Age older than 75, mild dementia, and nursing home residence are factors that negatively influence physicians' decisions to recommend mammography. Presence of chronic medical problems and functional limitations do not. Physicians appear to be using implicit judgments about quality of life and age rather than life expectancy based on comorbidity to determine breast cancer screening practices.