A comprehensive review of patterns of care studies from UCLA researcher Jennifer Malin and colleagues illustrates the difficulty of drawing conclusions about the quality of care, and “found marked variability in the patterns of breast cancer care both in the United States and other countries.”42

One reason for this may lie in the poor quality of cancer registry data as a source of information on outpatient treatments. For example, a study comparing patient records with data in the California Cancer Registry43 found that the registry correctly reported only 72 percent of radiation treatments, 56 percent of chemotherapy treatments, and 36 percent of tamoxifen use. By contrast, the accuracy of registry data was much higher for hospital-based procedures—95 percent for mastectomy and lumpectomy and 96 percent for lymph node dissection.

Studies examining the impact on mortality of access to treatment and utilization of treatment have yielded mixed findings and have usually been limited to the initial treatment of primary breast cancers, for example the four months that SEER data reflect. This points to an obvious need for better ways of looking at the entire dimension of treatment information in breast cancer.

There are at least five areas to consider regarding treatment, and they may often be interwoven. These are:

Consumer/patient beliefs and attitudes about medical treatment and providers can influence the other variables. Clearly, access to treatment—whether or not a woman can find the resources to receive standard-ofcare treatment for breast cancer—does not inevitably lead to treatment adherence and utilization, nor does it reflect the unspoken (and perhaps unconscious) bias that may exist in some health care providers leading to less than optimal treatment and poorer outcomes for patients who are older, obese, minorities—especially non- English-speakers—and those of lower socio-economic status. Any or all of these factors may interact to influence the actual treatment for breast cancer a patient receives—and the outcome of her disease, as a result. Not all women are able to take the time they need from work to travel for time-consuming treatments.

In an attempt to correct for these factors and examine whether or not a racial basis for differences in outcome exists, several studies have looked at outcomes for patients treated in single institutions, where financial and geographic barriers to access to care are removed or at least thought to be equivalent for different patient groups, and where all patients are offered the same treatments by the same doctors. An examination of survival data in NSABP cooperative group clinical trials found that when stage of disease and treatment were comparable, outcomes were similar for black and white breast cancer patients.44 However, a 10-year survival study done at M.D. Anderson Cancer Center found that even after correcting for the effects of age, socioeconomic status (SES), stage of disease, and delay in seeking treatment for symptoms, an ethnic difference remained among black, Hispanic, and white women.

A recently published study from the Center to Reduce Health Disparities45 found that even when correcting for staging, stage-specific six-year survival rates were lower for black women of all ages. But a closer examination of the data further defined the affected groups. It showed that “only black women younger than age 50 years with ER-positive tumors and women younger than age 65 with ER-negative tumors had significantly lower stagespecific survival rates.” The study also found that black women of all ages had lower rates of Stage I breast cancer. The study authors hypothesize that black women over the age of 65 may do better because of access to Medicare services, which may serve to alleviate disparities in cancer treatment.

Survival among Asian/Pacific Island (API) breast cancer patients has generally been found to be either equivalent to or superior overall to that of white women. A 2002 SEER-based study found different stage distributions and five-year outcomes for different subpopulations classified within the API group, finding higher stageadjusted survival rates for Japanese women, compared to Chinese and Filipinas. The study concluded that “The heterogeneity of cancer outcomes observed within the community classified as Asian reinforces the need for cancer statistics to be reported for disaggregated subgroups.”46

Other research points out that long-term compliance with adjuvant breast cancer treatments may be problematic in some patient groups, leading to poorer outcomes. For example, a recent study from the Journal of Clinical Oncology47 examined utilization of tamoxifen by tracking whether pharmaceutical prescriptions were refilled during the entire recommended five-year course of treatment. Tamoxifen use is important because it represents the most widely-used and effective treatment for hormone-positive breast cancer, representing the majority of cases.

Unlike more recently developed hormonal therapies (principally aromatase inhibitors), tamoxifen is no longer under patent, meaning that few economic barriers to its use persist. This study used prescription records of patients enrolled in New Jersey's Medicaid or Pharmaceutical Assistance to the Aged and Disabled (PAAD) programs. Analysis of these records over five and a half years found that almost one quarter of patients did not follow the five-year course of tamoxifen prescribed by their oncologists. Further, it found that nonadherence was highest among the youngest, oldest, and non-white patients.

The reasons for this are unknown, but this finding may well underscore the importance of an ongoing relationship with a medical team, an option not open to those with no insurance or means to pay for this level of care. For many people, hospital emergency rooms represent their only contact with health care.

A small study of patients treated for breast cancer in an HMO48 setting determined that missed appointments and later stage at diagnosis were “key determinants of survival.” After correcting for these variables, the effects of race were “marginal.” The study found that black women were more likely to miss their appointments, however, and concluded that “this measure is an important component of how race affects survival. Compliance with appointment keeping and alleviating reasons for noncompliance must be considered as factors in breast cancer survival.”

A study of Latinas in Massachusetts looked at the incidence of the attitude of fatalism as a predictor of poor screening behavior in minority women with family histories of breast cancer, and discussed cultural values in which open discussion of a cancer diagnosis was frowned upon, creating a potential for late diagnoses in familial cancers.49 Other studies have pointed at the low rates of routine regular medical care in minority populations and poor insurance coverage.

Surgery following neoadjuvant treatment for Stage III and surgery on the primary tumor for Stage IV patients may offer another insight into disparities in outcome. A 2000 Arkansas tumor-registry study50 on residual disease found that black women, who tend to be diagnosed at later stages, have “a lower incidence of becoming “disease-free” during treatment in Stage III and IV disease appears to be a contributing factor to the decreased survival in those stages.” The authors relate this to a higher rate of Stage IIIB disease combined with less surgical intervention among black women. More recently, research presented in the 2003 5th Annual Lynn Sage Breast Cancer Symposium on nearly 10,000 Stage IV patients in American College of Surgeons' National Cancer database demonstrated that survival is adversely affected when surgery is not performed on the primary tumor in metastatic breast cancer patients, or when surgery does not obtain tumor-free surgical margins.51

A study in older women, linking Medicare claims data to cases in the SEER database, showed that older women were less likely to receive radiation therapy after lumpectomy, a significant predictor of less favorable outcome. This failure to offer therapy was not found to be due to physician concern over other health issues. The authors concluded that “Between the ages of 65–69 years and 80 years or older, radiation therapy declined from 77 percent to 24 percent among women with no comorbid conditions.”52

Similar findings came from another large SEER-based study that found that black women of all ages were significantly less likely to receive radiation therapy after breast conservation.53

Data from the NCI Black-White Cancer Survival study54, comparing treatment received in these two populations with standard-of-care treatment in the mid-1980s, found that while only 4 percent of early-stage breast cancer patients were not treated with “minimum expected therapy,” that figure escalated to 36 percent of patients with late stage disease. “Older women and women with no usual source of care were significantly less likely to receive minimum expected therapy,” study authors reported. “Overall, 21 percent of black women did not receive minimum expected therapy compared to 15 percent of white women.”

A study of SEER Patterns of Care data from 1990––1998 in young breast cancer patients under the age of 35 found racial/ethnic disparities in both clinical characteristics and treatment.55 Young black women were less likely to have health insurance, had larger tumor sizes, more ER-negative tumors, and more positive lymph nodes, although fewer black women were offered axillary dissection. More black women and Hispanic women refused chemotherapy than white women. Both Hispanic and black women were less likely to receive radiation after lumpectomy, especially if they were on Medicaid. In this young population, combination chemotherapy was received by 46.5 percent of black women, 52.4 percent of Hispanic women, and 67 percent of white women. Not surprisingly, the study concludes, “African American and Hispanic women experienced poorer outcomes compared with white women. Just over 23 percent of African American and 21.9 percent of Hispanic women, compared with 15.6 % of non-Hispanic white women, had breast cancer listed as the underlying cause of death.”