© 2002 by Oxford University Press
Journal of the National Cancer Institute, Vol. 94, No. 5, 334-357,
March 6, 2002
© 2002 Oxford University Press
REVIEW |
Racial and Ethnic Disparities in the Receipt of Cancer Treatment
Affiliation of authors: Health Services and Economic Branch, Applied Research Program, Division of Cancer Control and Population Science, National Cancer Institute, Bethesda, MD.
Correspondence to: Vickie L. Shavers, Ph.D., National Institutes of Health, 6130 Executive Blvd., EPN 4005, MSC 7344, Bethesda, MD 20892–7344 (e-mail: shaversv{at}mail.nih.gov).
 |
ABSTRACT |
|---|
 Top Abstract BackgroundMethodsResultsDiscussionConclusionReferences |
|---|
A disproportionate number of cancer deaths occur among racial/ethnic minorities, particularly African Americans, who have a 33% higher risk of dying of cancer than whites. Although differences in incidence and stage of disease at diagnosis may contribute to racial disparities in mortality, evidence of racial disparities in the receipt of treatment of other chronic diseases raises questions about the possible role of inequities in the receipt of cancer treatment. To evaluate racial/ethnic disparities in the receipt of cancer treatment, we examined the published literature that addressed access/use of specific cancer treatment procedures, trends in patterns of use, or survival studies. We found evidence of racial disparities in receipt of definitive primary therapy, conservative therapy, and adjuvant therapy. These treatment differences could not be completely explained by racial/ethnic variation in clinically relevant factors. In many studies, these treatment differences were associated with an adverse impact on the health outcomes of racial/ethnic minorities, including more frequent recurrence, shorter disease-free survival, and higher mortality. Reducing the influence of nonclinical factors on the receipt of cancer treatment may, therefore, provide an important means of reducing racial/ethnic disparities in health. New data resources and improved study methodology are needed to better identify and quantify the full spectrum of nonclinical factors that contribute to the higher cancer mortality among racial/ethnic minorities and to develop strategies to facilitate receipt of appropriate cancer care for all patients.
|
BACKGROUND |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionConclusionReferences |
|---|
Approximately 1.2 million Americans were diagnosed with invasive cancer in the year 2001 (1). Racial/ethnic minorities are expected to account for a disproportionate number of these cancers. Overall age-adjusted cancer incidence is higher among African Americans than all other racial/ethnic groups (2). Cancer incidence rates are also higher among African Americans than whites for cancers of the prostate, lung and bronchus, colon and rectum, oral cavity and pharynx, cervix, and stomach. Similarly, cervical cancer incidence rates are higher among Hispanics than whites or African Americans. The age-adjusted incidence of stomach cancer for Asians/Pacific Islanders is more than twice the rate found among whites.
Approximately 552 000 Americans died of cancer in the year 2000 (1). A disproportionate number of these deaths occurred among racial/ethnic minorities, particularly African Americans, who have a 33% increased risk of dying of cancer compared with whites. Disproportionately high mortality rates are observed among African Americans for cancers of the lung, breast, prostate, colon and rectum, oral cavity and pharynx, cervix (2), and esophagus (3). In addition, the 5-year survival rates for African Americans are lower than those for whites for all of the major cancer sites (Table 1
).
|
In general, Hispanics have cancer mortality rates that are similar to non-Hispanic whites; however, rates for cervical and stomach cancers are higher among Hispanics than for non-Hispanic whites. Similarly, mortality rates for cancers of the cervix, liver, and stomach are higher among Asians/Pacific Islanders and Native Americans than among non-Hispanic whites. Overall breast cancer mortality rates for Native Hawaiians are the highest of all racial/ethnic groups (4). Nevertheless, African Americans are two times more likely to die of cancer than Asian/Pacific Islanders, American Indians, and Hispanics (5). Disparities in cancer mortality trends have also been noted. Mortality rates have been decreasing among both African Americans and whites overall, but decreases generally have been smaller and less consistent among African Americans than among whites (2), and racial disparities in mortality have persisted (2,6,7).
Racial/ethnic disparities in cancer outcomes are often attributed to a more advanced stage of disease at diagnosis among minorities (8–10). Differences in the stage of the cancer at diagnosis are thought to be primarily caused by the underutilization of cancer screening among racial/ethnic minorities (11–13). As a consequence, increasing access to screening has been the mainstay of the battle against racial/ethnic disparities in cancer mortality. Results from the National Health Interview Survey for 1987, 1992, and 1998 show that racial gaps in screening for breast and cervical cancers are closing for white, African-American, and Hispanic women; however, screening rates for colorectal cancer among African Americans and Hispanics continue to be substantially lower than rates for whites (14). There is also evidence of racial differentials in diagnostic follow-up of abnormal screening examinations (15–17) that might also contribute to higher disease stage at diagnosis among screened populations.
After adjusting for the stage of the cancer at diagnosis, however, several investigators (18–26) have reported residual disparities in cancer survival and mortality by race/ethnic group. This observation suggests that factors other than the stage at diagnosis contribute to the disparate cancer mortality observed among racial/ethnic minority populations.
Ideally, once diagnosed, patients will receive the appropriate treatment given their disease severity, prognostic factors, and comorbid conditions/or other contraindications to treatment. Studies of access to and/or utilization of noncancer medical services have, however, found racial differences in treatment after controlling for these factors. Race/ethnicity have been found to influence quality of care (27,28), service delivery (29), disposition after treatment in the emergency room (30), intensity of care provided to hospitalized patients (31), receipt of antiretroviral therapy (32,33), management of psychiatric disorders (34), receipt of certain cardiovascular procedures (29,30,35), and receipt of diagnostic cancer services (36). These findings coupled with the disproportionately higher cancer mortality of racial/ethnic minorities raise questions about possible disparities in the receipt of cancer treatment.
There are three broad categories of factors that have the potential to influence the receipt of optimal cancer care: structural barriers, factors that influence physician recommendations, and factors that influence patient freedom of choice and/or decision making. Structural barriers include insurance coverage and type and geographic location of the institution where care is to be received. Physician recommendations are influenced by factors, such as stage of disease, presence of certain prognostic indicators, perception of the willingness or ability of the patients to comply with treatment recommendations, preferences, and personal biases. Patient decision making is influenced by their attitudes and beliefs about specific treatments, their ability to navigate the medical system and to overcome structural and other barriers, and their personal preferences and biases (Fig. 1).
|
For this review, we define disparate treatment as that which is 1) a treatment that is not consistent with recommendations of the National Cancer Institute's Physician Data Query (PDQ) (37–42) and that has an adverse impact on outcome or 2) a recommended treatment associated with a poorer quality of life than another recommended alternative. The health professional version of the PDQ provides treatment summaries that include detailed information on prognosis, staging, and treatment of cancer. Potential disparities in the receipt of medical care can be identified by examining variations in utilization rates of specific procedures among patients with similar disease and other prognostic characteristics (43). Racial/ethnic patterns of care, therefore, are also described in this review for cancer sites for which PDQ lists treatment options but no preferred treatment.
To evaluate the existence of racial/ethnic disparities in the receipt of cancer treatment, we examined the published literature. We were interested in answering the following four questions: 1) Are there racial/ethnic variations in the receipt of cancer treatment? 2) If so, for which cancer diagnoses and types of treatment do these variations exist? 3) What is known about the specific mechanisms that explain these variations? 4) Do these treatment variations result in racial/ethnic disparities in cancer outcomes?
|
METHODS |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionConclusionReferences |
|---|
Several strategies were used to obtain the articles included in this review. We first searched MEDLINETM, CANCERLITTM, and HealthSTARTM databases for the period from 1990 through 2001 by use of the following keywords: lung cancer, breast cancer, colon cancer, cervical cancer, prostate cancer, cancer treatment, access to care, delivery of health care, health services accessibility, race, African American, black, Negro, Hispanic, Asian, Pacific Islander, Native American, American Indian, minorities, racial differences, efficacy, effectiveness, and underserved. We searched the literature for articles that provided data on efficacy/effectiveness, access/utilization, and trends in the use of specific therapeutic procedures. We also examined studies that focused on cancer survival or mortality by race/ethnic group with the belief that type of treatment was likely examined as an explanatory variable in multivariable analyses of these outcomes. We then examined the bibliographies of articles obtained from the first strategy. Finally, we conducted a search of articles in journals with a primary focus on minority health and of authors who had published articles on racial/ethnic patterns in access/utilization of cancer services. Each approach provided relevant articles. Only articles that describe research conducted in the United States and that describe racial/ethnic treatment patterns were reviewed. We excluded articles published before 1990, so that the provided information would describe current treatment patterns and practices. These articles provided data on patients treated from 1960 through 1997.
This review focuses on the racial/ethnic variation in treatment given to patients with breast, cervical, colorectal, prostate, or lung cancer. These cancers are estimated to have accounted for about 55% of all cancer incidence and more than 52% of all cancer mortality in 2000 (1). Well-defined treatment options exist for each of these cancer sites, although the relative efficacy of these treatments has not been definitively established by clinical evidence in all cases. In addition, the majority of published studies of cancer treatment patterns focus on these sites.
This review describes findings by cancer site. Each section begins with a description of stage-specific treatment options, which were obtained from The National Cancer Institute's PDQ (37–42). PDQ treatment summaries are peer reviewed and updated monthly by oncology specialists. Current literature on cancer treatment is reviewed from more than 70 biomedical journals. The section on treatment options is followed by one which provides results of articles on efficacy/effectiveness of specific treatments by race/ethnic group. Sections are also included on studies of trends in the patterns of care for specific treatments and studies that examine racial differences in access and utilization of specific treatments.
Both authors of this review independently reviewed all articles. Differences in the interpretation of findings were resolved by a consensus.
|
RESULTS |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionConclusionReferences |
|---|
Description of Studies
Eighty-seven studies were reviewed. Of these studies, 23 were clinical trials or equal treatment studies that provided data on treatment efficacy/effectiveness among racial/ethnic groups. We also reviewed eight studies that described treatment trends. The remaining 56 studies described racial/ethnic patterns of treatment. The majority of studies limited inclusion or analyses to African Americans and whites. Cancer-treatment patterns were described for Hispanics in 16 of the reviewed studies, for Asians in three, and for Native Americans in two.
Breast Cancer
Overview of Treatment Options
Clinical factors that influence breast cancer treatment selection include menopausal status, stage, axillary lymph node status, histologic and nuclear grade of the primary tumor, and estrogen receptor and progesterone receptor status. Specific PDQ recommendations for the treatment of breast cancer for patients not treated on a clinical trial protocol are provided in Table 2.
|
In 1990, the National Institutes of Health Consensus Panel on the treatment of early-stage breast cancer recommended breast-conserving surgery (BCS) with radiation therapy for stage I or II breast cancer (44). This treatment was recommended as preferable to mastectomy because of equivalent survival with preservation of breast tissue and potentially improved quality of life. BCS is only considered an equivalent option to modified radical mastectomy when it is followed by radiation. Observational research studies show that BCS without radiation is associated with higher recurrence rates than BCS with radiation (45,46). Since BCS with radiation and modified radical mastectomy result in equivalent survival for women with early-stage breast cancer, the decision to have BCS versus mastectomy is likely influenced by physician and patient preferences, including the desire for breast preservation and avoiding the inconvenience of multiple follow-up visits or side effects associated with radiation therapy.
As seen in Table 2
, systemic adjuvant chemotherapy or hormonal therapy is also recommended for women diagnosed with early-stage breast cancer, with specific types of adjuvant therapy depending on patient age, tumor size, nodal status and ER status.
Efficacy and Effectiveness We reviewed six studies that provided data on the efficacy and/or effectiveness of specific breast cancer treatment regimens by race/ethnic group. Roach et al. (47) reported that, after controlling for clinical prognostic factors, data from the Cancer and Leukemia Group B Trial, which began in 1985, show that African-American women, white women, and women of other races treated with adjuvant chemotherapy for stage II breast cancer had similar benefits. Data from the National Surgical and Adjuvant Breast and Bowel Project (NSABP) from 1981 through 1988 (48) and the Piedmont Oncology Association from 1974 through 1986 (22) also showed that African-American and white women with breast cancer received similar benefit from chemotherapy after surgical resection. A study comparing outcomes after breast-conserving surgery or mastectomy among African-American women treated at The University of Texas M. D. Anderson Cancer Center (Houston) from 1975 through 1994 revealed no differences in local recurrence, disease-free survival, and overall survival by type of treatment (49).
In contrast to these findings, Connor et al. (50) found higher systemic recurrence rates and shorter time to recurrence among African Americans than among whites treated from June 1982 through January 1998 with breast-conserving surgery and radiation therapy for stage II breast cancer at the University of Kansas Medical Center, Kansas City. There was, however, no statistically significant difference between African Americans and whites with regard to local recurrence rates for stage I or II breast cancer. In this study, African Americans treated from June 1982 through January 1998 were compared with white women treated from June 1982 through June 1990. In another study, Elias et al. (51) found that African-American women with stage I or II breast cancer had shorter overall survival and disease-free survival than white women who received similar treatment and had similar adjuvant chemotherapy compliance rates. Survival differences were confined to women with one to three positive lymph nodes.
There is also some evidence that suggests that the prevalence of estrogen receptor-positive breast tumors may be lower in African Americans and Hispanics than in whites (52,53), which might account for racial/ethnic differences in the use of tamoxifen. Furthermore, other evidence suggests that, because of the increased risk of stroke, pulmonary embolism, and deep vein thrombosis associated with tamoxifen, African Americans, who already have a higher prevalence of risk factors for these conditions, may receive less overall benefit from tamoxifen (54).
In general, however, there do not appear to be major racial/ethnic differences in the efficacy/effectiveness of breast cancer treatments. Among studies that did find differences, clinical factors that may have influenced the reported outcomes were often not examined in the analyses. For example, in studies by Elias et al. (51) and Connor et al. (50), comorbidity and other clinical prognostic indicators that might have explained the observed survival differences were not examined.
Trend Studies Since 1990, the use of breast-conserving surgery has increased among patients with early-stage breast cancer; however, mastectomy is still the most frequent therapeutic surgical procedure received (55,56). Data from the Detroit Cancer Surveillance System, a Surveillance, Epidemiology, and End Results (SEER)1 registry, show that use of breast-conserving surgery (57) increased from 5% from 1973 through 1977 to 42% from 1988 through 1992. Rates of breast-conserving surgery varied by age but were about the same for African Americans and whites. A more recent study (55) using data from the California Tumor Registry shows that, although breast-conserving surgery rates have increased among African-American women, they still lag behind utilization rates of similarly staged white women. An analysis of data from the SEER registries shows that the use of breast-conserving surgery for early-stage cancers ranges from a low of 26.7% in Iowa to a high of 55.6% in Connecticut (58). Receipt of recommended adjuvant therapy has also been shown to be declining among some rural African-American and white women (56).
Access/Utilization Studies Breast-conserving surgery. Several studies (59–61) found racial variation in the receipt of breast-conserving surgery. Data from the Black/White Cancer Survival Study for 1985 and 1986 show that, after controlling for tumor size and comorbidity, African Americans were less likely than whites to receive breast-conserving surgery and radiation therapy instead of mastectomy (60). Medicare claims data for 1986 (62) and 1990 (61) indicated that elderly African Americans were significantly less likely than whites to receive breast-conserving surgery.
In contrast, analysis of SEER data from 1983 through 1986 showed no racial/ethnic differences in the receipt of breast-conserving surgery (63), as did a study of treatment for early-stage breast cancer diagnosed from 1985 through 1987 conducted by the Metropolitan Detroit Cancer Surveillance System (64). Although Ballard-Barbash et al. (65) found that the receipt of breast-conserving surgery was similar for African Americans and whites in SEER areas from 1985 through 1989, rates of breast-conserving surgery were somewhat lower for other nonwhite patients with breast cancer.
More recent studies (56,59,66,67), however, have demonstrated few differences between African Americans and whites in the receipt of appropriate surgical treatment of breast cancer among treated patients. Tropman et al. (56) reported that rural African-American and white women diagnosed with stage I or II breast cancer received breast-conserving surgery at a similar frequency in 1991 and 1996, after adjusting for comorbidity, estrogen receptor status, tumor size, and clinical trial enrollment. Data from the California Cancer Registry from 1988 through 1995 (55) and from SEER–Medicare from 1988 through 1993 (68) show that the rate of breast-conserving surgery was higher among African Americans than among whites and other racial/ethnic groups examined. In contrast, data from the Carolina Breast Cancer Study from 1995 through 2000 showed that fewer African Americans than whites received breast-conserving surgery (59).
Racial/ethnic differences in the receipt of breast-conserving surgery seem to be more prevalent among other nonwhite populations, particularly Asians and Hispanics, than among African Americans. Morris et al. (55) noted that Asian and Hispanic women received breast-conserving surgery at a statistically significantly lower frequency than white women. Similarly, Lee et al. (69) found that, among women diagnosed from January 1990 through December 1992, Chinese women were less likely than African-American and white women to receive breast-conserving surgery and reconstructive surgery after mastectomy.
Adjuvant radiation therapy. Racial/ethnic disparity in breast cancer treatment is more evident in the receipt of adjuvant radiation therapy after breast-conserving surgery. Analysis of SEER data from 1983 through 1986 shows that African Americans were less likely than whites to receive radiation therapy after breast-conserving surgery (63), as did an analysis of SEER data from 1985 through 1989 (65) and SEER data from 1988 through 1993 (68). Similarly, data from a convenience sample of hospitals showed that, among elderly women diagnosed from November 1995 through September 1997, African Americans were two times more likely than whites not to receive radiation therapy after breast-conserving surgery (70).
Other studies. Several other studies of patterns of breast cancer care during the 1980s provide some evidence of racial variations in the receipt of breast cancer treatment. Included are differences in the receipt of appropriate diagnostic testing, definitive treatment, conservative surgery, and follow-up after diagnosis and initial treatment. When screened by the National Breast and Cervical Cancer Early Detection Program from July 1, 1991, through September 30, 1995, racial/ethnic minority women had longer periods from the first abnormal screening to diagnosis and from diagnosis to treatment initiation than white women (71). Data from the University of San Francisco Mobile Mammography Screening Program for women screened from July 1993 through May 1994 also show that nonwhite women have longer follow-up times after an abnormal mammogram than white women (15). In crude analyses of data from the Black-White Cancer Survival study for 1985 and 1986, African-American women were less likely than white women to receive minimum expected therapy (72). Racial differences in receipt of minimum expected therapy, however, were not statistically significant when data were stratified by stage at diagnosis (early/late). In addition, an analysis of SEER–Medicare data for women diagnosed in 1991 showed that African Americans were significantly less likely than whites to have a mammogram after their diagnosis and treatment of breast cancer (73).
Studies reporting no racial/ethnic minority treatment disparities. Several studies reported no racial/ethnic disparities in treatment patterns. Although PDQ-recommended adjuvant chemotherapy and hormonal therapy were less frequently received than primary therapy (i.e., surgery and radiation therapy), Tropman et al. (56) found no racial differences in the receipt of these treatments between rural African-American and white women in data from the Reaching Communities for Cancer Care Project in 1991 or 1996. African-American, Hispanic, and Asian patients with breast cancer in selected California counties diagnosed from 1984 through 1990 were statistically significantly more likely than white patients to receive chemotherapy and less likely to receive hormone therapy as treatment of local and regional-stage breast cancer (24). Data from the New Mexico Tumor Registry for women diagnosed with breast cancer for 1994 and 1995 showed that racial/ethnic minorities were more likely than whites to receive radiation therapy after breast-conserving surgery (74). Siminoff et al. (75) found no difference between whites and nonwhites in the rate of referral to oncologists after surgical treatment of breast cancer in a population-based sample of data collected from 1993 through 1995.
Summary for breast cancer. We reviewed 23 studies that examined racial/ethnic patterns of care for breast cancer. Fourteen of these studies provided data on treatment patterns only, and eight provided data on both treatment and survival. Twelve of the 23 studies found racial variations in treatment (22,24,52, 59–61,65,68,70–73). Of these 12 studies, six found statistically significant differences (24,52,60,68,71,73). Two of the three studies (24,52) that reported statistically significant racial/ethnic treatment variations and provided survival data reported survival disparities.
The four largest population-based studies (24,61,65,68), with the greatest statistical power, analyzed data from 1985 through 1993. All studies performed multivariable analyses that adjusted for clinical and nonclinical factors, which varied between studies. Overall, results from these studies were mixed. There does, however, seem to be some consistency in the findings among studies that examined the receipt of radiation therapy after breast-conserving surgery, suggesting that African Americans less frequently receive radiation therapy after breast-conserving surgery (65,68).
Cervical Cancer
Overview of treatment options.
The selection of a specific treatment for stage IA1–IA2 cervical cancer may be influenced by the desire for ovarian preservation, comorbid conditions, and potential late side effects. PDQ-recommended stage-specific treatment options for cervical cancer are provided in Table 2.
Efficacy and effectiveness.
No racial differences in outcomes were found among patients treated with radical hysterectomy for stage IB and IIA cervical cancers (76) (Table 3). African-American, Hispanic, and white women had similar response rates to treatment with chemotherapy for advanced cervical cancer (77). Postoperative radiation therapy provided similar survival for African-American and white women who had pathologic risk factors for recurrence (78). In contrast, Han et al. (79) found that African Americans treated with brachytherapy and external-beam radiation therapy had statistically significantly poorer survival than whites for stage II cervical cancer only. Analyses for this study did not control for the prevalence of comorbidity at baseline. The authors suggest that differences in patients with stage II disease may have contributed to the observed survival disparities.
|
Trend studies. The use of radiation therapy alone or in conjunction with other treatment decreased from the early 1980s to 1990 (80,81), while the use of hysterectomy alone increased (81). These patterns were observed for whites, Hispanics, and African Americans; however, declines in the use of radiation therapy alone as treatment of cervical cancer are less pronounced for African Americans than for other race/ethnic groups (81). This result was thought to be associated with the higher proportion of African-American women diagnosed at more advanced disease stages, where radiation therapy alone is more frequently used.
Access/utilization studies. Racial/ethnic disparities in the receipt of treatment of cervical cancer were found in the receipt of clinical staging, any definitive treatment, surgery, and intracavitary irradiation. An analysis of SEER data from 1988 through 1994 showed that African Americans were more likely than whites to go untreated and to have radiation therapy alone and were less likely to have surgery alone as treatment of their cervical cancer after controlling for stage and age (82). Similarly, another analysis of SEER data from 1992 through 1996 (83) found that African Americans were less likely than whites to receive clinical staging, and unstaged patients were less likely to be treated. SEER data from 1990 through 1995 indicate that African-American women diagnosed with stage 1A1–1A2 cervical cancer and Hispanic women with stage 1A2 cervical cancer were more likely than white women with similar disease to be treated with the less adequate fertility-sparing procedures and less likely to be treated with hysterectomy. Among women over the age of 35 years, African Americans were still more likely than white women to receive fertility-sparing procedures than hysterectomy (84). African-American women with International Federation of Gynecology and Obstetrics stage IB cervical cancer were more likely than white women to be treated with radiation therapy alone and less likely to be treated with surgery or combined therapy (85). African Americans with cervical cancer less frequently received intracavitary radiation therapy as opposed to external-beam radiation therapy than white women (86).
Summary for cervical cancer. We reviewed five studies with information on racial patterns of cervical cancer care (82–86). Three of these studies were population based (82–84), and all three found racial variations in treatment. The two remaining studies analyzed data from medical records of single institutions. A study conducted at Grady Memorial Hospital (Atlanta, GA) found racial variations in cervical cancer treatment (86), whereas the study conducted at the University of Chicago Hospitals (IL) did not (85). In addition to racial treatment variations, Howell et al. (82) also found survival disparities as did Mundt et al. (86); however, findings in Mundt et al. were not statistically significant.
Findings from the limited evidence provided by the reviewed studies suggest that African Americans are less frequently treated for their cervical cancers (82,83). Some evidence also suggests that, when treated, African Americans may more frequently receive fertility-sparing treatment for stage 1A2 cervical cancer (83), which is not a PDQ-recommended treatment option for such patients (see Table 2).
Some factors that influenced racial/ethnic variations in treatment included differences in the prevalence of comorbid conditions (83,86), poorer health (86), patients' refusal of treatment (83,86), and lack of physician recommendation for treatment (83).
Colorectal Cancer
Overview of treatment options.
Surgical resection is recommended for all stages of colorectal cancer. Because failure rates differ by site (colon or rectum), adjuvant therapy recommendations differ by site. PDQ stage-specific treatment recommendations for cancers of the colon and rectum are provided in Table 2.
Efficacy and effectiveness. Trials conducted by the NSABP demonstrated similar efficacy for African Americans and whites for five adjuvant therapeutic regimens for colorectal cancer (87). No statistically significant differences in tumor response or disease progression were found among African Americans and whites with colon cancer enrolled in the Department of Veterans Affairs (VA) Cooperative Study 188 Chemotherapy Trial (88).
Trend studies. Use of anal sphincter-sparing surgery for the treatment of rectal cancer has been increasing, as has multimodal treatment of colon cancer (89).
Access/utilization studies. Racial/ethnic variations were found in the receipt of major therapeutic procedures, including surgical resection (89–91). African Americans were less likely than whites with similar disease to receive a major colorectal cancer therapeutic procedure (90), cancer-directed surgery (89, 91), and sphincter-sparing surgery (89). African Americans also were less likely than whites to receive adjuvant therapy for stage III colon cancer (92,93) and resection for advanced colon cancer (94). Hispanics were less likely than non-Hispanics to receive chemotherapy (95). Although several studies found that racial/ethnic minority patients with colorectal cancer were less likely to receive endoscopic follow-up after potential curative surgical treatment (96–98), their results were not statistically significant.
In contrast, data from the National Cancer Data Base showed few racial/ethnic differences in the type of treatment received (99). African Americans and Native Americans in this study, however, were more likely than other racial and ethnic groups to be untreated for their colon cancer.
Few racial/ethnic differences were found in treatment of colorectal cancer among patients treated in equal-access systems, where payment and referral patterns are not a barrier to receipt of treatment. For example, African Americans and whites treated for colorectal cancer at VA hospitals had similar rates of surgical resection, radiation therapy (88,100), and chemotherapy (100).
Only one study reported that racial/ethnic minorities more frequently received a specific treatment for colorectal cancer. In contrast to other study findings (89–91,95), Roetzheim et al. (95) found that African Americans treated for colorectal cancer in Florida were more likely than whites with a similar stage disease to receive definitive surgery for their colorectal cancer.
Summary for colorectal cancer. The reviewed evidence on racial/ethnic differences in treatment of colorectal cancer is not totally consistent. Studies that did not find differences tended to be nonpopulation based, small studies of a single locality, or studies from institutions characterized as equal-access systems. In contrast, larger population-based studies (91,93,95) and national surveys (89) more frequently found differences in treatment patterns by racial/ethnic status. Four of the five population-based studies showed statistically significant racial/ethnic differences in the receipt of cancer treatment from 1984 through 1996. Evidence from the studies that show differences indicate a consistent pattern of the receipt of less aggressive treatment among nonwhites than among whites, including receipt of any colorectal cancer-directed treatment (89–91,94,95), adjuvant therapy (92,93), and follow-up after initial potentially curative treatment (96,97).
Lung Cancer
We reviewed five studies that provided data on racial/ethnic patterns of treatment of lung cancer and three studies of efficacy/effectiveness. Treatment of lung cancer has remained relatively stable, and so we were unable to find studies that examined changes in treatment trends. Many new therapeutic regimens are still in clinical trials. The lack of any major changes in treatment coupled with the high fatality rate associated with lung cancer contributes to PDQ recommendations to enroll patients in clinical trials.
Overview of Treatment Options
Small-cell lung cancer.
Small-cell lung cancer (SCLC) can seldom be surgically resected at the time of diagnosis (101) because patients tend to present with disseminated disease. Neither surgery nor radiation therapy provides long-term survival. Addition of chemotherapy to the treatment regimen increases median survival fourfold to fivefold (42). Because SCLC tends to metastasize to the brain, prophylactic cranial irradiation is often recommended (Table 2). The designation of limited or extensive disease is used in lieu of tumor–node–metastasis staging (178) to describe the extent of disease among patients with SCLC.
Non-small-cell lung cancer.
Approximately one third of non-small-cell lung cancers (NSCLCs) can be surgically resected at the time of diagnosis (101). PDQ stage-specific treatment recommendations for SCLC and NSCLC are provided separately in Table 2.
Efficacy and Effectiveness No statistically significant differences in tumor response or disease progression were found among African Americans and whites diagnosed with NSCLC or SCLC enrolled in the VA Cooperative Study 188 Chemotherapy Trial (88). Similarly, African Americans and whites treated with surgical resection for NSCLC experienced similar 5-year survival (102). Graham et al. (103) also found that African Americans and whites with NSCLC had similar survival after receiving radiation therapy.
Access/Utilization Studies African Americans less frequently received surgical resection (88,102,104–106), radiation therapy (88), chemotherapy (107), or any definitive treatment (106) for their lung cancer.
SEER–Medicare data from 1985 through 1993 show that African Americans were significantly less likely than whites to receive surgical resection for stage I and stage II NSCLC (102). Similarly, African Americans entering a chemotherapy trial from 1981 through 1986 were, at the time of trial enrollment, statistically significantly less likely than whites to have received surgical resection or radiation therapy for their lung cancer (88). African Americans were more likely than whites to have had metastases at the time of diagnosis. SEER data for Detroit (MI), San Francisco (CA), and Seattle–Puget Sound (WA) from 1978 through 1982 (104) show that, among patients with early-stage NSCLC, whites were 20% more likely than African Americans to receive surgical resection of their lung cancer. In addition to race, median family income was also a predictor of the receipt of surgical resection. SEER–Medicare data from 1991 through 1993 show that African Americans were the least likely of all other racial/ethnic groups to receive chemotherapy for metastatic NSCLC (107). In spite of having the same insurance, African Americans were about 30% less likely than whites to receive palliative chemotherapy. Other factors that negatively influenced the receipt of chemotherapy for metastatic NSCLC were lower socioeconomic status and treatment in a nonteaching hospital. SEER data from 1988 through 1995 indicate that, among patients diagnosed with NSCLC, African-American patients less frequently received a recommendation for surgery than whites (108). Similarly, an analysis of Medicare data from 1985 through 1989 showed lower surgical resection rates for local/regional stage NSCLC and higher rates of no definitive treatment of distant-stage NSCLC among African Americans than among whites; however, these differences were not statistically significant (105). Data from the National Cancer Data Base for 1986 and 1987 indicate that African Americans were less likely to receive treatment for NSCLC than all other racial/ethnic groups examined (106).
Summary for Lung Cancer African Americans were consistently found to less frequently receive treatment for their lung cancers and, when treated, to not receive surgery. Evidence from the reviewed studies suggests that racial differentials in the receipt of surgical resection have persisted throughout the period covered from 1978 through 1995 by the reviewed studies. All of the reviewed studies analyzed data from population-based databases, had large sample sizes, and provided evidence of racial/ethnic variation in the receipt of treatment of lung cancer. Results in one, however, were not statistically significant, and the majority reported survival disparities (102,104,105,108). In two of the studies, all patients also had Medicare insurance coverage (102,107), which should have reduced disparities in access to care.
Prostate Cancer
Overview of treatment options. Several factors should be considered in deciding appropriate treatment options for patients diagnosed with prostate cancer. These factors include stage, histologic grade, age, comorbid conditions, prostate-specific antigen (PSA) levels, and site of metastases (38).
Because of the many side effects associated with the various treatments, the definition of appropriate prostate cancer therapy remains somewhat controversial (109,110). The side effects associated with radical prostatectomy include urinary incontinence, urethral stricture, and impotence. Data from a case series from 1994 through 1997 suggest that new nerve-sparing surgical techniques have lower rates of incontinence and higher rates of potency preservation (111). Complications associated with the use of external-beam radiation therapy include acute cystitis, proctitis, and enteritis. Although potency may be initially preserved, it often diminishes over time. The side effects associated with the use of hormonal therapy include loss of libido, impotence, hot flashes, gynecomastia, and psychologic effects. Stage-specific treatment recommendations for prostate cancer from PDQ are provided in Table 2.
Efficacy and effectiveness. Of the 17 studies that focused on efficacy and effectiveness, 11 concluded that race/ethnicity were not independent predictors of treatment failure. Race/ethnicity were not predictive of poorer outcomes after treatment with radiation therapy (112–115), radical prostatectomy (113,116–118), hormone therapy (119,120), or chemotherapy. Disease progression (121) and risk of secondary treatment after initial management with "watchful waiting" were found to be similar between African Americans and whites (122). Disease-specific survival did not differ by race for patients treated with radiation therapy (117,123) or for patients with negative surgical margins (118). Incidence-based mortality was found not to differ statistically significantly between African Americans and whites after controlling for stage, grade, age, number of primary cancers, and treatment (124). Iselin et al. (118) found no statistically significant racial differences in cancer survival or time to increased PSA levels of patients treated with prostatectomy when stratified by Gleason grade or extent of disease. Similarly, after adjusting for stage and grade, Austin et al. (115) found no statistically significant racial differences in cause-specific survival after radiation therapy for prostate cancer. Results from Witte et al. (125) indicate that race/ethnicity are not independent predictors of positive surgical margins after radical prostatectomy.
Several other studies, however, have concluded that race/ethnicity did influence treatment outcomes. African Americans with palpable tumors were found to have statistically significantly lower median times to biochemical failure (PSA elevation and disease-specific survival among patients with positive surgical margins after radical prostatectomy) (118). Among patients treated with radiation therapy, African Americans were found to have a larger proportion of tumors with post-treatment PSA levels greater than 10 ng/mL and lower overall survival (126). After controlling for stage, however, statistically significant racial differences in overall and cause-specific survival were evident for stage C disease only. When the analysis was controlled for grade, racial differences were noted for overall but not cause-specific survival. In another study, African Americans had poorer survival after adjusting for stage and grade (127). In crude analyses, the risk of death for African Americans who received prostatectomy was more than four times higher than that for whites (128). Additional analyses revealed that the increased risk of death was limited to the first 12 months after diagnosis. Although not specifically examined, comorbidity was not believed to be a major contributor to racial disparities in overall survival because there were no statistically significant differences in the cause of death among deceased participants. Furthermore, Shekarriz et al. (129) concluded that race was an independent prognostic factor for disease-free survival among men who had tumors with positive surgical margins.
Racial disparities in treatment outcomes in these studies were believed to be associated with unmeasured cultural (128) and biologic (118,128) differences. Among studies that found racial disparities in survival in crude analyses, few statistically significant survival differences remained after controlling for stage and/or tumor grade (115,126).
Trend studies. Radical prostatectomy rates have been increasing among both whites and nonwhites (130). From 1973 through 1988 and from 1989 through 1996, Shaw et al. (131) found similar trends between African Americans and whites in the use of prostatectomy. The rate of radiation therapy was similar for African Americans and whites from 1973 through 1988 but was significantly higher for African Americans than whites from 1989 through 1996.
Access/utilization studies. Data from the reviewed studies suggest that whites more frequently receive aggressive treatment of prostate cancer than African Americans (25,109,132–140). African Americans were statistically significantly more likely than whites to go untreated (25,109,128,133), less likely to receive definitive treatment (140) and radical prostatectomy (25,109,132–134,136), and more likely to receive conservative management (watchful waiting) for prostate cancer (134). Native Americans and Hispanics were more likely than whites to go untreated and, when treated, to receive prostatectomy (128). An analysis of data from the California Cancer Registry found that treatment received by African Americans was more frequently in concordance with the PDQ recommendations than treatment received by whites because African Americans less frequently received prostatectomy, which is discouraged for older men with advanced disease (132).
In contrast, data from the Connecticut SEER registry from 1988 through 1992 showed that the receipt of prostatectomy for local/regional stage and endocrine surgery for distant-stage prostate cancer are similar among African-American and white patients with prostate cancer (141). In another study, data from the North Carolina Central Cancer Registry for 1994 and 1995 showed that African Americans were more than three times more likely than whites to have prostatectomy for distant-stage disease (142). An analysis of data in the Virginia Cancer Registry from 1985 through 1989 showed that African Americans were more frequently treated with orchiectomy than were whites (109).
Summary for prostate cancer. Prostate cancer is notable among cancer sites in that there is considerable evidence, albeit from the minority of published studies on the topic, that race may be an independent factor in prognosis after equivalent treatment. However, it is not clear whether this observation is caused by underlying behavioral, biologic, or genetic factors associated with race or because of incomplete or inaccurate measurement of clinical risk or quality-of-care factors. Furthermore, it is also not clear that systematic differences in treatment patterns can be explained by inherent differences in treatment outcomes. Just as there is a paucity of definitive evidence on the relative efficacy of the various treatment approaches for prostate cancer, there is also insufficient evidence to conclude that any one treatment modality is more preferable for African Americans than for other racial groups.
In general, relative to whites, African Americans and other minorities are less likely to receive the relatively expensive or more innovative treatments. Almost all published observational studies of treatment patterns, including several large, multisite, population-based studies (25,132,134,138,139), have found that, for any given time period, African Americans are less likely than whites to receive radical prostatectomy and more likely to receive conservative management, although there has been an increasing trend in radical prostatectomy for African Americans over time. Similarly, African Americans were more likely to receive orchiectomy rather than the newer and more expensive hormonal drug treatment. Generally, treatment differences were not explained by differences in clinical factors. For example, in analysis of data from the National Cancer Data Base, Mettlin et al. (136) concluded that, despite little variation in disease stage and tumor grade, prostate cancer treatment varied by age, race, and region. However, comorbidity was not controlled for in all studies and may not have been controlled for completely in any of the studies. Given the uncertainty about the comparative efficacy and quality-of-life effects associated with these competing treatment modalities, without more definitive clinical evidence than is now available, it cannot be concluded that African-American patients, on average, experience less effective or less appropriate treatment for prostate cancer than whites.
As noted with colorectal cancer, fewer racial/ethnic differences were found among patients treated in equal-access systems. After controlling for stage at diagnosis, there were no statistically significant differences between African Americans and whites with regard to prostate cancer treatment received by patients in the Department of Defense Tumor Registry, an equal-access treatment facility (8).
|
DISCUSSION |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionConclusionReferences |
|---|
Two dimensions of treatment have been hypothesized to contribute to racial disparities in cancer-treatment outcomes (6). The first is racial differences in treatment efficacy, and the second is the failure to provide suitable care. Our review provided very limited evidence of racial/ethnic differences in treatment efficacy or effectiveness among patients with similar disease characteristics (e.g., stage, grade, or comorbidities). The majority of studies included in this review, however, reported racial disparities in cancer treatment, including receipt of a definitive primary therapy, adjuvant therapy, conservative surgery, and follow-up after potentially curative treatment (Tables 3 and 4
).
|
The receipt of treatment can be influenced by type and/or teaching status of the hospital (62,64,91,145). Racial/ethnic variation in referral patterns or use of specific institutions (146–149), therefore, may contribute to treatment disparities observed in multi-institutional studies. We also note that fewer racial/ethnic treatment disparities were found in studies of single institutions and equal-access systems (66,67,79,143,144).
Several nonclinical factors were implicated as contributors to racial/ethnic disparities in the receipt of cancer treatment. These include socioeconomic status (55,61,72), patient concerns about body image (70), age at diagnosis (72), the size of the hospital where surgery is performed, the number of previous surgeries performed by the surgeon (64), geographic location (63), and insurance coverage (72). Patient preferences and decisions also contributed to treatment differences (82,83,86). For example, Merrill et al. (83) found that 7.4% of African-American women who did not receive cancer-directed surgery for cervical cancer refused it.
The reviewed studies also suggest that white patients receive more aggressive therapy than other racial/ethnic groups for lung, prostate, and colorectal cancers, even in cases where evidence for clinical benefit is less than definitive. For example, whites more frequently received inappropriately aggressive treatment for stage I colorectal cancer (94) and advanced prostate cancer (132,137). Conversely, whites are more likely to receive less aggressive therapy in cases where the clinical benefit of a less aggressive regimen is well established, for example, breast-conserving surgery (55,59,61).
Equal Treatment Equal Outcomes?
Studies of clinical trials and equal-access systems provide support for the idea that equal treatment will yield similar cancer outcomes between racial/ethnic minorities and whites with similar disease (47,67,77,87,88,103,150,151). Additional support is provided by studies that examine the survival experiences of similarly staged and treated cancer patients (66,76,78,118,122, 123,152,153). Racial/ethnic minority and white patients who receive similar care for the same stage of disease have been shown to have similar survival experiences for multiple myeloma (152) and cancers of the cervix (76–78), lung (88,103), colon (87,88), and prostate (118,122,123,150,151,153). Data from the VA demonstrate that comparable cancer outcomes can be obtained among patients who receive similar treatment irrespective of socioeconomic status (88).
Limitations of the Current Literature
The current studies were limited in a number of ways, including insufficient sample sizes and the failure to control for relevant clinical factors in multivariable analyses. Twenty-three percent of the reviewed studies had fewer than 100 participants per race/ethnic group studied, and 13.2% had fewer than 200 participants per race/ethnic group studied. It is worthwhile to note that these studies less frequently reported statistically significant racial/ethnic treatment differences than studies with larger sample sizes. Comorbid conditions can influence both the type of treatment (154) and survival (150,154). Racial/ethnic minorities have a higher prevalence of comorbid conditions (154), which emphasizes the importance of evaluating the role of comorbidity in studies examining racial/ethnic differences in receipt of cancer treatment, particularly, adjuvant therapy.
Few of the studies were able to examine the role of patient preferences and patient–physician decision making in treatment patterns, especially where there are trade-offs between treatment effectiveness and treatment side effects. Similarly, few of the studies were able to examine the specific mechanisms that resulted in the observed disparities in treatment process or outcomes. Pathologic staging data were sometimes compared with clinical staging data. Patients are frequently upstaged after surgical resection (155); thus, the use of pathologic stage for the retrospective examination of treatment differences may not provide an accurate reflection of the decisions made on pretreatment clinical presentation, especially with regard to surgical resection rates.
Census tract or ZIP Code level data were often used to measure socioeconomic status, which may not be highly correlated with individual level data (156). In addition, racial/ethnic differentials in underreporting in census data (157) suggest that the reliability of these data may differ among racial/ethnic groups. Furthermore, it is not known whether socioeconomic status has the same relative meaning for different racial/ethnic groups (158). Racial bias modifies the influence of education and income so that racial/ethnic variations may exist within the same income and educational levels. Arbes et al. (159) argues that, because race represents a social construct and socioeconomic status is a consequence of race, socioeconomic status should not be modeled as a confounder because it would bias the hazard ratio for race toward the null. In contrast, others have found that race and socioeconomic status have independent effects on cancer mortality (160), survival (161), and receipt of medical services (162).
This review is not the first to note these limitations (163). Recognizing the need to obtain more informative data on health disparities in cancer, the NCI has initiated a number of efforts to expand its research base in this area (164). These efforts include expansion of SEER registries to improve coverage of racial, ethnic, and other groups who suffer from disproportionately high cancer mortality rates and linkage of these registries to census level and individual level data on socioeconomic status, the development of a cancer outcomes research group designed to study cancer processes of care and outcomes in large populations spanning diverse population groups and treatment settings (165), and the establishment of centers for population health designed to expand our understanding of social and environmental causes of cancer-related health disparities.
Future Research Directions and Study Design Considerations
The manner in which race/ethnicity and other nonclinical factors influence the receipt of cancer treatment is complex, as Fig. 1 shows. Whether racial/ethnic differences in receipt of cancer treatment noted in the reviewed studies reflect differing health-care needs, preferences, racial bias, or other intervening factors is not clear. Additional studies are needed to provide information on understudied racial/ethnic groups (i.e., Asians, Pacific Islanders, and Native Americans) and to examine the influence of access to care, socioeconomic status, and clinical staging, as well as beliefs and attitudes of patients and providers and their influence on racial/ethnic disparities in cancer treatment.
Access to care. Racial/ethnic differentials in the prevalence of barriers may influence the receipt of optimal cancer care (166,167). In a study of cost barriers to cancer treatment, African Americans more frequently than both Hispanics and whites reported that they lost their medical insurance coverage after their diagnosis, were denied coverage after changing jobs, and reached their insurance spending limits (167).
Socioeconomic status. The complex interaction between race/ethnicity and socioeconomic status makes it difficult to disentangle the independent effects of these two variables and may overshadow more proximal determinants of treatment. Therefore, knowledge about the influence of specific social or economically influenced barriers to cancer treatment, such as transportation to medical appointments, type of health insurance, type of health-care facility and provider, copayment amounts, availability for care (i.e., the ability to take time off work, availability of child care), support systems, knowledge of appropriate care, and attitudes toward cancer treatment may prove more informative for the development of interventions designed to reduce health disparities.
Clinical staging. Studies included in this review suggest that racial/ethnic minorities more frequently than whites have unknown/unstaged cancers (83,168), which influences whether they are treated (83) and/or which treatments are selected (169). It is not apparent whether the racial variations in clinical staging are caused by the refusal of patients to be staged, the failure of the clinician to perform clinical staging, or the failure to record stage in the medical record.
Patient's beliefs and attitudes. Health beliefs and attitudes influence health-care behavior (170). The prevalence of fatalistic and nihilistic attitudes is higher among racial/ethnic minority populations than white populations (171). Higher treatment refusal rates among racial/ethnic minorities (82,83,86), therefore, may reflect differences in attitudes toward treatment. Additional research is needed to further elucidate the origins of and role of patient preferences and decision making in racial/ethnic cancer treatment disparities.
Provider beliefs and attitudes. Findings from several studies suggest that a physician's perception of patients may be influenced by nonclinical characteristics (172–174), which may be manifested in differences in patient referral patterns and treatment recommendations (172,174,175). In a survey of eight New York State hospitals, physicians were found to more frequently have negative perceptions of African Americans and persons of low or middle socioeconomic status than of whites and persons of high socioeconomic status (173). This finding and the fact that there is little information on how physician attitudes toward patients impact their care suggest the need for further research, particularly with regard to how such perceptions might contribute to racial/ethnic disparities in cancer treatment.
Biologic factors/tumor aggressiveness. Racial/ethnic differences in prevalence of adverse biologic characteristics and tumor aggressiveness may also contribute to disparities in cancer outcomes (52,79,126,133). Elledge et al. (52) found that, compared with white women, African American women more frequently have less favorable prognostic factors, such as fewer estrogen receptor- and progesterone receptor-positive tumors and higher S-phase fractions. Chen et al. (176), however, found that African Americans with colon cancer experienced poorer survival, despite less aggressive tumors. Other studies did not find statistically significant racial/ethnic differences in molecular prognostic indicators for breast cancer (177) or the burden or growth rate of androgen-independent prostate cancer (120). Therefore, it is important to determine the role, if any, that differences in tumor biology and other clinical factors play in racial/ethnic disparities in cancer treatment and outcomes.
|
CONCLUSION |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionConclusionReferences |
|---|
In this review, racial/ethnic minorities were found to have received disparate cancer treatment in several studies. Racial/ethnic minorities were consistently found to have less frequently received appropriate surgical resection for lung and colorectal cancers. Although some racial/ethnic variations in surgical resection were also noted for breast and prostate cancers, study findings were not conclusive. Racial/ethnic disparities were also consistently found in the receipt of a cancer-directed therapy, radiation therapy after breast-conserving surgery, clinical staging, and adjuvant therapy. In many cases, racial/ethnic disparities in treatment were not explained by differences in clinical profiles.
Reducing the influence of nonclinical factors on the receipt of cancer treatment may provide an important means of reducing racial/ethnic disparities in health. New data resources and improved study methodology are needed to better identify and quantify the full spectrum of nonclinical factors that contribute to the disparate mortality of racial/ethnic minority populations and to develop strategies to facilitate appropriate cancer care for all patients.
|
NOTES |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionConclusionReferences |
|---|
1 Editor's note: SEER is a set of geographically defined, population-based, central cancer registries in the United States, operated by local nonprofit organizations under contract to the National Cancer Institute (NCI). Registry data are submitted electronically without personal identifiers to the NCI on a biannual basis, and the NCI makes the data available to the public for scientific research.
We thank Drs. Rachel-Ballard-Barbash, Robert Hiatt, and Jon Kerner from the NCI, Bethesda, MD, for their comments and suggestions on various drafts of this manuscript.
|
REFERENCES |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionConclusionReferences |
|---|
1 American Cancer Society. Cancer facts & figures; 2001.
2 Ries LA, Eisner MP, Kosary CL, Hankey BF, Miller BA, Clegg L, Edwards BK, editors. SEER cancer statistics review, 1973–1998. National Cancer Institute (Bethesda, MD); 2001.
3 Farrow DC, Vaughan TL. Determinants of survival following the diagnosis of esophageal adenocarcinoma (United States). Cancer Causes Control 1996;7:322–7.[CrossRef][ISI][Medline]
4 Shinagawa SM. The excess burden of breast carcinoma in minority and medically underserved communities. Cancer 2000;88(5 Suppl):1217–23.[CrossRef][ISI][Medline]
5 Greenlee RT, Murray T, Bolden S, Wingo PA. Cancer statistics, 2000. CA Cancer J Clin 2000;50:7–33.[Abstract]
6 Dignam JJ. Differences in breast cancer prognosis among African-American and Caucasian women. CA Cancer J Clin 2000;50:50–64.[Abstract]
7
Chu KC, Tarone RE, Brawley OW. Breast cancer trends of black women compared with white women. Arch Fam Med 1999;8:521–8.
8 Optenberg SA, Thompson IM, Friedrichs P, Wojcik B, Stein CR, Kramer B. Race, treatment, and long-term survival from prostate cancer in an equal-access medical care delivery system. JAMA 1995;274:1599–605.[Abstract]
9
Eley JW, Hill HA, Chen VW, Austin DF, Wesley MN, Muss HB, et al. Racial differences in survival from breast cancer: Results of the National Cancer Institute Black/White Cancer Survival Study. J Natl Cancer Inst 1995;87:1686–93.
10 Liu WL, Kasl S, Flannery JT, Lindo A, Dubrow R. The accuracy of prostate cancer staging in a population-based tumor registry and its impact on the black-white stage difference. Cancer Causes Control 1995;6:425–30.[CrossRef][ISI][Medline]
11 Suarez L. Comparing acculturation scales and their relationship to cancer screening among older Mexican-American women. J Natl Cancer Inst Monogr 1995;18:41–7.
12 Hegarty V, Burchett BM, Gold DT, Cohen HJ. Racial differences in use of cancer prevention services among older Americans. J Am Geriatr Soc 2000;48:735–40.[ISI][Medline]
13 Sung HY, Kearney KA, Miller M, Kinney W, Sawaya GF, Hiatt RA. Papanicolaou smear history and diagnosis of invasive cervical carcinoma among members of a large prepaid health plan. Cancer 2000;88:2283–9.[CrossRef][ISI][Medline]
14 Breen N, Wagener D, Brown ML, Davis WW, Ballard-Barbash R. Progress in cancer screening over a decade results of cancer screening from the 1987, 1992 and 1998 NHIS. National Health Interview Surveys. J Natl Cancer Inst 93:1704–13.
15 Chang SW, Kerlikowske K, Napoles-Springer A, Posner SF, Sickles EA, Perez-Stable EJ. Racial differences in timeliness of follow-up after abnormal screening mammography. Cancer 1996;78:1395–402.[CrossRef][ISI][Medline]
16 McCarthy BD, Yood MU, Boohaker EA, Ward RE, Rebner M, Johnson CC. Inadequate follow-up of abnormal mammograms. Am J Prev Med 1996;12:282–8.[ISI][Medline]
17 Carey P, Gjerdingen DK. Follow-up of abnormal Papanicolaou smears among women of different races. J Fam Pract 1993;37:583–7.[ISI][Medline]
18 Cooper GS, Yuan Z, Rimm AA. Racial disparity in the incidence and case-fatality of colorectal cancer: Analysis of 329 United States Counties. Cancer Epidemiol Biomarkers Prev 1997:6:283–5.[Abstract]
19 Vernon SW, Tilley BC, Neale AV, Steinfeldt L. Ethnicity, survival, and delay in seeking treatment for symptoms of breast cancer. Cancer 1985;55:1563–71.[CrossRef][ISI][Medline]