Compliance rates for diagnostic mammography and the effect of race on breast density are highlighted in two studies in the April issue of the American Journal of Roentgenology. While Michigan researchers looked at the costs associated with a system designed to encourage patients to return for additional imaging, a group in Massachusetts sought to determine if innate racial differences in breast density might explain differences in breast cancer risk.
16¢ per patient
Locking in 100% compliance for additional tests for their onsite mammography screening program was not difficult for the breast imaging faculty at the University of Michigan in Ann Arbor, as diagnostic exams could be ordered while the patient waited for screening results, wrote Dr. Caroline Blane and colleagues. But the group wanted to maintain that high standard at an offsite screening program as well.
In addition, they wanted to "document the hidden additional costs of achieving better compliance with patient recall from a batch-read screening mammography program," the group explained (AJR, April 2007, Vol. 188:4, pp. 894-896).
The scheduling process for batch reading included informing the patient that a recall for a diagnostic study may be required. All patients placed in final BI-RADS categories 3-5 automatically had a diagnostic study. However, those given a BI-RADS score of 0 were contacted by telephone within two days of screening for additional imaging. Letters were also sent to the referring physician and the patient (the latter received a registered letter). The authors pointed out that they did not rely on the referring physician to act on the letter.
The researchers prospectively measured the clerical time in 100 consecutive recalls and calculated the average cost at $17 per hour (salary plus benefits).
According to the results, during a three-year period, 13% of the 30,286 batch-read screening patients were BI-RADS 0 and recalled for more imaging. Of these 4,025 women, 98.8% returned for a diagnostic study after being contacted by phone. An average of 2.2 phone calls (3.64 minutes of total clerical time) was recorded for 100 patients. The clerical cost per case was calculated at $1.03.
Of the 4,025, 1.2% was noncompliant and received an average of six phone calls (4.7 minutes of time) for a clerical cost of $1.33 per case. But after receiving the registered letter, 58% of these women returned for more imaging, and one cancer was detected.
The overall compliance rate was 99.5%, with the total additional cost of 16¢ per screening patient, the authors reported. They noted that the key to achieving high compliance was to systematically track noncompliant patients, including using a radiology information system (RIS) that searches for a specific incomplete exam code and alerts the clerical staff. In turn, the staff must pull the charts to review the case. Finally, scheduling clerks are trained so that any patient who indicates a problem (lump, history of breast cancer) will automatically receive an onsite diagnostic exam.
A 'suitable surrogate' for cancer risk?
Breast density has been seen as an independent predictor of cancer risk and, on mammography, dense tissue makes it difficult to detect cancers. In addition, women of color have a lower incidence of cancer than Caucasian women, but generally present with more advanced disease. Dr. Marcela del Carmen and colleagues investigated whether "mammographic density would explain differences in breast cancer risk among racial groups" (AJR, April 2007, Vol. 188:4, pp. 1147-1150).
This retrospective study was done at the Avon Foundation Comprehensive Breast Evaluation Center at Massachusetts General Hospital in Boston and included 15,292 women. Of this population, 83.1% were Caucasian, 10% were other (such as Native American), 3.67% were African-American, and 2.84% were Asian.
Mammographic density was categorized using BI-RADS, while body mass index (BMI) was calculated using the National Institutes of Health (NIH) definition (25-29.9 kg/m² representing overweight). BI-RADS frequency counts were calculated across the BMI categories. By filling out a questionnaire, the women self-reported race, age, bra size, cup size, weight, and height.
According to the authors' analysis, Asian women had dense breasts (BI-RADS 4), while the trend in African-American women was toward lower density. Caucasian women had higher breast density than African-American women and lower than Asian women, the group wrote.
"Age and BMI predict density with a 73.4% concordance level.... Adding a term for Asian race ... while controlling for BMI and age improved the prediction of breast density significantly (73.5% concordance)," the authors stated. "The breast density among Asians was documented to be statistically greater.... However, in all of the other groups, breast density did not correlate with race beyond what can be attributed to differences in age and BMI."
Instead, other factors -- bra size, cup size, and BMI -- were more likely to account for the any racial differences in breast density, they added. As a result, racial groups with lower breast density, and a lower cancer risk, did not have lower mammographic breast density.
"If breast density is related to breast cancer risk, either Asian women should have a higher incidence of breast cancer or they should have the least dense breasts," del Carmen's group explained. Instead, their study showed that Asian women had the highest breast density but a lower incidence than Caucasian women. They concluded that mammographic density was not a suitable surrogate of cancer risk differences across racial groups.
By Shalmali Pal
AuntMinnie.com staff writer
March 30, 2007
Related Reading
Mammography results poorly communicated to blacks, February 23, 2007
Mammographic density an independent risk factor for breast cancer, January 18, 2007
Minority women still face lag time for breast cancer diagnosis, treatment, December 27, 2006
Breast density may improve breast cancer prediction models, September 6, 2006
Copyright © 2007 AuntMinnie.com
