The Lynx Group

Reasons for Breast Cancer Care Disparities Remain Unclear

July/August 2010, Vol 1, No 3

Chicago, IL—Insurance status and socioeconomic status (SES) explain little of the racial and ethnic variation in the receipt of recommended therapy for breast cancer, found Rachel Freedman, MD, MPH, in her examination of a national cancer registry.

She sought to determine the variables that might explain differences in care between 662,117 white (86% of the cohort), black (10%), and Hispanic (4%) women diagnosed with a first invasive breast cancer during 1998-2005 at hospitals included in theNational Cancer Database, a hospital-based cancer registry run by the American College of Surgeons and the American Cancer Society. Fifty-one percent had private insurance and 40% had Medicare.

For this analysis, 4 variables were assessed:

  • Receipt of the locoregional therapy
  • Receipt of hormonal therapy in the adjuvant setting
  • Receipt of chemotherapy in the adjuvant setting
  • Hormone receptor testing at diagnosis.

At about 98%, “hormone receptor testing was very high for all groups, which is reassuring, but there were subtle differences for all of the other measures by race, particularly the black patients versus the white patients,” said Dr Freedman, a medical oncologist at Dana-Farber Cancer Institute, and instructor in medicine at Harvard Medical School, Boston.

Definitive locoregional therapy was instituted for 74.3% of Hispanic women, 75.5% of black women, and 80.6% of white women. Receipt of adjuvant hormonal therapy was 40.2%, 45.4%, and 54.4% for Hispanic, black, and white women, respectively; and receipt of adjuvant chemotherapy was 59.5%, 58.4%, and 48.7% for Hispanic, black, and white women, respectively.

“Accounting for insurance barely attenuated any of the differences, so although there were insurance differences, race is definitely more important than insurance in this data set,” Dr Freedman emphasized.

Socioeconomic status was also not associated with outcome but “the SES data we had were area level and not individualized SES,” she said.

Large Data Set, Big Limitations

“These big data set projects have tons of limitations because we don’t have any information about reasoning for omission of treatment, provider information, or patient stresses. Although we controlled for some prognostic variables, we didn’t have information on HER-2 positivity and other things that can be a part of why people receive care. There were also comorbidities that we couldn’t account for.”

Other studies have shown that minorities are more likely to mistrust their doctors and less likely to enroll in clinical trials, and whether such differences in behavior are responsible for differences in receipt of treatment by race is not known, said Dr Freedman.

“What this study shows is that fixing insurance and expansion of insurance in the United States may not fix everything, and that there are other things that we have to tackle,” she concluded.

She added that examination of the receipt of treatment by patient age is forthcoming.

“We really can’t look much further unless we want to look at hospital characteristics such as volume and quality, which is one of the things I’m trying to do,” she said.

An assessment of other large databases, such as the National Comprehensive Cancer Network, may provide more clues about factors that explain disparities in receipt of recommended breast cancer care, she said.

Related Articles

Subscribe to
Value-Based Cancer Care

Stay up to date with personalized medicine by subscribing to recieve the free VBCC print publication or weekly e‑Newsletter.

I'd like to recieve: