This population-based study of 2,777 Medicare breast cancer survivors who underwent breast cancer surgery in 2003 found both racial and SES predictors of treatment in a low-volume hospital. Those treated at low-volume hospitals were more likely to be black or Hispanic, to live in a rural setting, and to be of a lower SES. We also found that the greater distance they had to travel to a high-volume hospital, the less likely they were to seek care from these hospitals. The distances traveled to high-volume hospitals (≥40 cases/year) were greater for women with a lower household income who lived in rural areas, who were less educated, who were married, and who had less emotional support. Nonetheless, racial and socioeconomic disparities in treatment at a low-volume hospital persisted even after controlling for distance from residence to the nearest high-volume hospital. Being black or Hispanic, living in a rural area, having a lower household income, and living a greater distance from the nearest high-volume hospital were all factors independently associated with treatment at a low-volume hospital.
Breast cancer is the cancer with the highest incidence among women in the United States.32
We have shown previously that the vast majority of this care is decentralized, with only 10% of patients treated by surgeons who are performing at least 30 operations per year.33
Recognizing this decentralization of care, interest has grown in the hospital volume–breast cancer outcome relationship. Prior studies have demonstrated improved survival with hospitals that treat higher volumes of breast cancer patients.12,15,21,23
The nature of this relationship is complex and the specific mechanism is unclear.
Given these findings, treatment in low-volume hospitals would appear to be less desirable, particularly for those patients who are ethnic minorities or who are from a low SES—two groups that are particularly vulnerable to health care disparities. Low-volume hospitals have been shown to offer lower rates of radiation treatments and offer less definitive treatment.13,14,24
These institutions may have less access to specialized providers, state-of-the-art treatments, or clinical trials, which may be related to a poor outcome. Their ability to provide multidisciplinary, coordinated care may also be limited, in addition to their ability to assess their own quality of care.
Beginning in 2009, the New York State Department of Health set forth a policy that mandated that Medicaid beneficiaries receive breast cancer surgery services at hospitals and ambulatory surgery centers that perform at least 30 of these procedures annually. Facilities that perform less than this threshold will not be reimbursed for these procedures provided to Medicaid patients. This policy is consistent with existing evidence supporting better breast cancer outcomes with higher case volumes.34
Prior research that used New York State Cancer Registry data and the New York State hospital discharge database demonstrated that patients operated on at high-volume hospitals had far better 5-year survival, similar to our own work on a broader group of hospitals.15
In France, a decree from the Ministry of Health and Solidarity in 2007 was issued that requires hospitals treating breast cancer to demonstrate that they perform ≥30 breast cancer surgeries a year.35
However, it will be important to demonstrate whether these policy changes will actually affect breast cancer outcomes in their respective geographical areas.
These changes may not produce the desired effect if patients are unwilling to travel to high-volume centers. Reluctance to travel has been previously shown.36,37
We found that poorer black and Hispanic patients from rural areas were more likely to go to low-volume hospitals. Centralizing high-quality care may further polarize existing disparities. Requiring patients to seek care from high-volume centers will increase travel distances, as we have shown here, and as others have also shown.38
Also, patients may not obtain care as promptly if they are forced to travel longer distances. However, this policy change may be successful in places like New York and France, where an established infrastructure exists for public transportation.
There are several limitations to this study. First, this study focuses solely on Medicare patients and may not be generalizable to the younger breast cancer population. However, the fact that all patients were Medicare beneficiaries was a strength of the study in that it should mitigate differences in insurance coverage and patient access to high-volume centers. We found that despite these factors, racial and socioeconomic factors persisted.
Another limitation is that we may have some inaccuracy in hospital distance measurements. For patients residing near a state border, the nearest hospital could be located in another state. We only measured distances to in-state facilities. However, demonstrates that there is homogeneity in distances to nearest hospitals, and distances to the nearest high-volume hospitals are consistent with population density (that is, subjects who lived in rural areas lived a greater distance from a high-volume hospital).
Finally, although we found that minority and low-SES patients are more likely to be operated on at low-volume hospitals, we cannot conclude that this finding is the sole reason for survival disparities in breast cancer. This finding is likely one of several factors that govern survival. However, we can infer that this factor may well partially account for the survival disparities.
With this study, we have established that in this population-based cohort of Medicare breast cancer patients, those who are black or Hispanic, who live in rural areas with lower incomes, and who live farther from a high-volume hospital are more likely to be treated at a low-volume hospital. Public policies may help change some of these discrepancies, such as those currently instituted in New York and France. However, it remains important to assess whether these regulations positively affect outcomes. The mechanism that drives the volume–outcome relationship in breast cancer survival has yet to be fully elucidated, as does the threshold of procedures or patients treated, which influences outcomes. In addition, high-volume status is an imperfect indicator of better outcomes. Future studies will need to examine the influence of processes of care and factors at the hospital level to explain differences in survival.