African-American and white head and neck carcinoma patients in a university medical center setting
Are treatments provided and are outcomes similar or disparate?
Racial and ethnic disparities occur in many areas of the health care management system in the United States. These disparities include disease incidence, access to health and medical services, treatments provided, and disease outcomes. Health care delivery organizations have limited resources. Encounters between patients and providers in health care delivery organizations typically are cross-cultural. Access to care, quality of care, and equity may be affected by limited resources and cross-cultural encounters. This impacts the diagnosis, treatments provided, and outcomes, with African-American patients faring poorly compared with white patients. African Americans are 15% more likely to develop cancer than whites and are about 34% more likely to die of cancer than whites in the United States. The purpose of this study was to determine and compare the characteristics of African-American patients and white patients with carcinoma of the head and neck at the University of Cincinnati Medical Center, an equal-access facility, reporting similarities and disparities in disease stage at the time of diagnosis, treatment received, and patient outcomes. Cancer 2001;91:279–83. © 2001 American Cancer Society.
The hospital records and tumor registry abstracts of 54 African-American patients and 52 white patients who were diagnosed and treated for carcinoma of the head and neck at the University of Cincinnati Medical Center between 1991 and 1996 were selected randomly from a larger sample and reviewed retrospectively. Forty-four of the African-American patients were men, and 33 of the white patients were men. The mean age at diagnosis was 59.8 years for African Americans (range, 44–81 years) and 59.3 years for whites (range, 39–87). Criteria for inclusion in this study required an original occurrence, a pathologic diagnosis of squamous cell carcinoma, diagnosis and treatment between 1991 and 1996, an approximately equal number of African-American and white patients, and the availability of American Joint Committee on Cancer/International Union Against Cancer classification and stage grouping of malignant tumors.1 Follow-up records for these patients were reviewed through December 31, 1999. Nine patients were lost to follow-up.
In addition to medical, surgical, and pathologic information, nonmedical data were collected, including insurance, marital status, education, occupation, tobacco and alcohol use, and other factors. Some of this information was not available in the records that were reviewed. Death certificate information was obtained for cause of death. Patient survival was calculated from the time of pathologic diagnosis to either death or the most recent contact.
MATERIALS AND METHODS
Cancer and Demographics
In 1999, the Institute of Medicine Committee on Cancer Research Among Minorities and the Medically Underserved published a report entitled The Unequal Burden of Cancer: An Assessment of National Institutes of Health Research and Programs for Ethnic Minorities and the Medically Underserved.2 Noting that “the burden of cancer in specific populations is a critical tool in the establishment of research priorities,” the report described the cancer burden experienced by many ethnic minorities in the United States. African-American men and women have the highest reported mortality rates from all cancers (except for female lung carcinoma) of all racial/ethnic populations in the United States. African-American men have the highest rates of prostate carcinoma in the world. Lung carcinoma incidence and mortality rates for African-American men are 53.5% and 45% higher, respectively, than the incidence rates for white American men. Although the incident rates for carcinoma of the colon and rectum in African Americans declined between 1990 and 1995, they are still higher than in any other racial and ethnic group. The death rates for colon and rectal carcinoma among African Americans are about 30% higher than those in other racial and ethnic groups.3 For oral cavity and pharyngeal disease sites, the mortality rate for African Americans has been almost twice the mortality of whites (5.4 vs. 2.9) The highest cancer mortality rate was for African-American men (9.5).4
There is a correlation between tobacco and alcohol use and head and neck carcinoma. Tobacco is the most preventable cause of premature death in the United States and is responsible for about 30% of all cancer deaths. Recent national data suggest that the prevalence of smoking among African-American youth has increased. From 1991 to 1997, cigarette smoking among African-American male students doubled from 14.1% to 28.2%.5 In this study population, 100% of African American patients and 88% of white patients reported significant smoking history, and more than 80% of all patients reported alcohol use.
Each patient was assigned to one of nine head and neck carcinoma sites. There were no patients with nasopharyneal or subglottic tumors in this study group. The most common sites for African Americans were supraglottic, glottic, oropharynx, and oral cavity. For whites, the four most common sites were oropharynx, oral cavity, supraglottic, and glottic. For the purposes of analysis, the original nine sites were condensed into larynx, pharynx, sinus, and oral cavity (Tables 1 and 2).
Table 1. Descriptive Statistics of 54 African-American Patients and 52 White Patients with Head and Neck Carcinoma who were Diagnosed and Treated Between 1991 and 1996
| I||4 (15)||3 (18)|
| II||4 (15)||4 (23)|
| III||12 (44)||8 (23)|
| IV||7 (26)||2 (12)|
| Total||27 (100)||17 (100)|
| I||0 (0)||4 (18)|
| II||3 (18)||4 (18)|
| III||4 (23)||9 (41)|
| IV||10 (59)||5 (23)|
| Total||17 (100)||22 (100)|
|Carcinoma of the oral cavity|
| I||3 (30)||2 (15)|
| II||2 (20)||4 (31)|
| III||2 (20)||4 (31)|
| IV||3 (30)||3 (23)|
| Total||10 (100)||13 (100)|
Table 2. TNM Classification by Race
The management of patients with carcinoma of the head and neck is complex and may involve single-modality treatment for approximately 40% of patients who present with Stage I or II disease. Combined-modality therapy is recommended for patients with locally advanced disease.6 In this study group, surgery was considered minimum treatment. Ninety-eight percent of all patients in the group underwent surgery. Of the African-American patients who underwent surgery, 26% underwent surgery alone, 68% underwent surgery and received radiation therapy, 4% underwent surgery and received both radiation therapy and chemotherapy, and 2% underwent surgery and received chemotherapy. Of the white patients who underwent surgery, 49% underwent surgery alone, 43% underwent surgery and received radiation therapy, and 8% underwent surgery and received both radiation therapy and chemotherapy.
For patients with head and neck tumors, disease stage at the time of diagnosis is the single factor that is most predictive of survival.6 In this study group, for patients with Stage I–III disease, African-American patients and white patients were similar, with eight more cases reported for whites in these stages. African-American patients presented with Stage IV disease twice as often as white patients (20 patients vs. 10 patients, respectively). Laryngeal tumors represented 43% of all cases, pharyngeal tumors represented 36%, and oral cavity tumors represented 21%. Six sinus carcinoma patients were included in the pharyngeal carcinoma category.
The goal of this study was to determine whether all patients received comparable treatment regardless of their cultural background. We found that 98% of all patients received at least surgical treatment. Of the African-American patients who underwent surgery, 26% underwent surgery alone, 68% underwent surgery and radiation therapy, 4% underwent surgery and received both radiation therapy and chemotherapy, and 2% underwent surgery and received chemotherapy. Of the white patients who underwent surgery, 49% underwent surgery alone, 43% underwent surgery and received radiation therapy, and 8% underwent surgery and received both radiation therapy and chemotherapy (Table 3). Twice as many African-American patients presented with Stage IV disease. The average death rate over time was about twice as great for African-American patients. There were 51 deaths in this study group through the follow-up date, December 31, 1999: 33 African-American patients and 18 white patients. Deaths among African Americans occurred most often (75%) from disease specific causes. Among white patients, death from disease specific causes occurred 25% of the time (Table 4).
Table 3. Number and Percent of Patients Receiving Each Type of Treatment by Race
|Surgery||14 (36)||25 (64)||39 (100)|
|Surgery and radiation||36 (62)||22 (38)||58 (100)|
|Surgery, radiation, and chemotherapy||2 (33)||4 (67)||6 (100)|
|Radiation||1 (100)||0||1 (100)|
|Surgery and chemotherapy||1 (100)||0||1 (100)|
|Radiation and chemotherapy||0||1||1 (100)|
Table 4. Number and Percent of Deaths Through December 31, 1999, by Cause
|Total deaths||33 (65)||18 (35)||51|
|Disease specific causes||18 (75)||6 (25)||24|
|All causes||15 (56)||12 (44)||27|
Using logistic regression, the odds ratio (OR) at 95% confidence intervals was calculated relating death to race, with adjustment for age at diagnosis, treatment site, and disease stage. The results revealed a significantly greater risk of death among African-American patients compared with white patients (P < 0.05), as shown in Table 5. The OR of 2.8 indicates that African-American patients have an almost three-fold greater risk of death at the end of the follow-up period compared with white patients. Using the Cox proportional hazards regression model (Table 6) reveals the conditional risk ratios at 95% confidence intervals and relates the number of years to death (or outcome) to race with adjustment for age at diagnosis, treatment, tumor site, and disease stage. This analysis revealed a significantly greater rate of death for African-American patients compared with white patients (P < 0.01). In addition, a risk ratio of 1.5 indicates that later stage disease increases the death rate by 50% after adjustment for age at the time of diagnosis, type of treatment, tumor site, and disease stage at diagnosis (P = 0.02). A risk ratio of 2.2 indicates that African-American patients have an approximately two-fold greater rate of death compared with white patients in this study group. The mean survival from the time of diagnosis to death was 2.1 years for both African-American patients and white patients.
Table 5. Odds Ratios and 95% Confidence Intervals Relating Death at the End of Follow-Up to Race with Adjustment for Age, Treatment, Tumor Site, and Disease Stage
|Race (African American vs. white)||2.8a||1.2–6.7|
|Age at diagnosis (5 yrs)||1.1||0.9–1.3|
|Surgery vs. other treatmentb||0.8||0.2–3.4|
|Surgery and radiation vs. other treatmentb||0.8||0.2–3.3|
|Larynx vs. other sitec||0.8||0.2–2.3|
|Pharynx vs. other sitec||0.9||0.3–2.8|
Table 6. Conditional Risk Ratios and 95% Confidence Intervals for the Model Relating Number of Years to Death (or Censoring) to Race with Adjustment for Age, Treatment, Tumor Site, and Disease Stage
|Race (African American vs. white)||2.2a||1.1–4.2|
|Age at diagnosis (1 yr)||1.0||0.9–1.0|
|Surgery vs. other treatmentb||0.9||0.3–2.7|
|Surgery and radiation vs. other treatmentb||0.6||0.2–1.8|
|Larynx vs. other sitec||0.7||0.4–1.5|
|Pharynx vs. other sitec||1.0||0.5–2.1|
The hypothesis that, in an equal-access facility, patients with head and neck carcinoma receive equal minimum treatment was verified in this study. Ninety-eight percent of the patients received the single modality, surgery. In the combined-modality group, 60% were African-American patients, and 40% were white patients. African-American patients presented with Stage IV disease twice as often and died twice as often as white patients. The unequal burden of cancer among African Americans in this study population in relation to more advanced disease and poorer outcome has been demonstrated. What, then, may be possible sociologic factors that affect later disease stage at the time of presentation and poorer outcomes for African Americans? These may include the following: 1) Is there access to health care services? Even if services are available, is there a range of services available, are hours feasible, and is transportation available? 2) Does the patient have a primary care physician? 3) What (if any) type of insurance does the patient have? 4) Is there an awareness level regarding risk factors for head and neck cancer? 5) If there is a primary care physician, will that individual refer to secondary and tertiary resources? 6) Does cancer fatalism or fear in knowing the truth result in delay in seeking care? 7) Is there variability in how disease progresses between racial groups?
These factors need additional study and analysis and may lead to increased funding for research, health education programs, community awareness campaigns, service availability in neighborhoods, and factors that allow all citizens to utilize health care services easily. Hopefully, this study has provided information and knowledge that will lead to changes in health care policy and funding that can reduce the unequal burden of cancer among minorities and the medically underserved.
At the University of Cincinnati Medical Center, an equal-access medical facility, 98% of all patients in this study group underwent surgery. Sixty-one percent (65 patients) received combined-modality therapy. Of that group, 60% were African-American patients, and 40% were white patients. The higher mortality outcomes and later disease stage at the time of presentation for African Americans may be the result of sociologic factors, such as access to health care, insurance, awareness of risk factors, and variability in disease progression. Insurance information for this study group will be analyzed and reported later. Other sociologic factors warrant further study and analysis.
The authors acknowledge Shanna Armstrong, R.N., Director of the Cancer Registry at the University of Cincinnati Medical Center, Barrett Center, who provided assistance in obtaining Cancer Registry data and patient follow-up information. The authors also acknowledge Linda Levin, Ph.D., of the University of Cincinnati Medical Center, Center of Biostatistical Services, who assisted with the interpretation and analysis of data.