CHICAGO—Cancer patients who are medically underserved or who receive different care because of ethnic/racial disparities suffer the consequences, including higher risk for many cancers, low rates of screening, later stage of disease at diagnosis, fewer treatments or treatments not adherent to guidelines, disease progression, and high morbidity and mortality relative to the general population. These problems were the focus at a Clinical Science Symposium here at the ASCO Annual Meeting.
“These disparities in care present a complex problem that includes symptom recognition, ability to pay for care, and language barriers in communication. There are differences in comorbid conditions that can mask symptoms, exacerbate cancer, and be a counter-indication for treatment,” said the Discussant for the session, Cathy Bradley, PhD, Chair of Healthcare Policy and Research at Virginia Commonwealth University Massey Cancer Center.
“Patients of lower socioeconomic status may see providers who are less well trained. Individuals with lower incomes may have a narrow spectrum of specialists to choose from who will take uninsured or Medicaid patients. These are problems that can involve multiple factors that can influence whether an individual patient has disparities in treatment and cancer outcomes.”
High Symptom Levels in Medically Underserved
Charles Cleeland, PhD, Professor of Cancer Research, at the University of Texas MD Anderson Cancer Center, noted that under-management of cancer-related symptoms, including pain, fatigue, and nausea, persist even in the best centers, and there is concern that the medically underserved are at even greater risk for high pain levels and poor pain management, as well as fatigue, disturbed sleep, and distress.
Dr. Cleeland led a longitudinal study of advanced lung cancer patients from one tertiary cancer center and three public hospitals treating underserved patients.
He and his colleagues performed a multi-symptom assessment of 13 common cancer symptoms before and during chemotherapy for up to 15 weeks using the patient-rated MD Anderson Symptom Inventory. They also examined symptoms that interfered with normal function.
“We used a computer telephone-based assessment tool that called patients weekly and the patients rated their symptoms. We took the most severe symptoms to form a composite score and classified patients as having consistently high or low symptom burden,” he said.
The study enrolled 185 patients with Stage IIIB or IV non-small cell lung cancer, 102 patients from a tertiary care cancer center in Houston, and 83 patients from three public hospitals, two in Houston and one in Miami. All patients completed at least baseline pre-chemotherapy assessments; the overall median number of assessments was 10. Some 107 patients completed assessments for up to 15 weeks.
Patients at the tertiary cancer care center had more severe disease, and there were no differences in the combination therapy, single-agent therapy, or targeted therapies received.
“The major finding is that we identified two clearly different symptomatic groups—one with clinically significant symptoms,” Dr. Cleeland said. The two predictors of the high-symptom group were poor performance status and being treated in a public hospital.
Only 25% of the high-symptom group came from the tertiary cancer care center; in the public hospitals, the high-risk patients ranged from 45% to 62%.
“Patients in the public hospitals with less severe disease survived longer,” he said. Under-management of pain was found in 26% of those in the tertiary cancer care center compared with 50% of those in the public hospitals.
“The level of symptom distress can vary greatly, even in patients with the same cancer treatments. In this example, treatment at one tertiary cancer care center was associated with markedly lower symptom distress compared with public hospitals. We saw this very early in treatment. There were no corrections in symptom management across time.”
The implication, he said, is that similar practices in cancer treatment do not assure similar symptom management. “Higher symptom levels in the medically underserved are likely to be determined by care, patient resources, and patient adherence.”
Dr. Cleeland's study had a small sample size but good measures of symptom burden, Dr. Brady said in discussing the study. “It's unclear where the source of disparity originates: Is it with the providers in these public hospitals, or contextual factors that lead to the higher risk and lower levels of treatment?
“We have found in our studies that patients who are underserved tend to miss more appointments than any other group of patients. Is that related to their toxicity or the difficulty in their lives to come in for treatment?”
Dr. Cleeland might be able to randomize patients to specific protocols and facilities, but “can an intervention that works in one setting be transferred to another setting? With different patients come different challenges,” Dr. Brady said.
Racial Disparities in Prostate Cancer
Another study reported the symposium, led by Ronald Chen, MD, Assistant Professor in the Department of Radiation Oncology at the University of North Carolina-Chapel Hill, examined prostate cancer rates in African American and Caucasian men receiving treatments in concordance with published guidelines.
Dr. Chen used National Comprehensive Cancer Network (NCCN) guidelines and Surveillance, Epidemiology and End Results (SEER) tumor registry data linked with Medicare health care claims data for health care services received and comorbid conditions.
Included were 13,725 patients (2,174 African American and 11,551 Caucasian) diagnosed in 2004-2005 with localized prostate cancer Stage T1-3a, non-metastatic, who had no other cancer diagnoses. The researchers examined whether the patients received treatment within 12 months of diagnosis and in concordance with NCCN guidelines.
Significantly more Caucasians (66%) received treatment in concordance with NCCN guidelines than African Americans (60%). On the other hand, significantly more Caucasians (50%) were likely to have high-risk disease compared with African Americans (43%).
“Most of the discordance was due to receiving no treatment within 12 months or androgen-deprivation therapy alone,” said Dr. Chen. In a multivariate analysis, race, increasing age, comorbidity, and high-risk disease were all associated with low odds of receiving concordant treatments.
In conclusion, Dr. Chen said, “The differences may contribute to our understanding of racial disparities in prostate cancer treatment outcomes. The modest differences between African American and Caucasian men in receipt of guidelines-concordant treatment is unlikely to fully explain the large outcome differences seen in other studies.”
As a next step, Dr. Brady suggested that Dr. Chen make more use of the data he has collected to look at geographic regions, control for census track characteristics, and get provider information and do a statistical analysis of interactions between race and risk groups.
“Does being African American put you in a higher risk group, or is there a synergistic effect of being African American and high-risk disease?” she asked.
These types of disparities are difficult to resolve because of inadequate data and inadequate interpretation of that data. “We are trying to address a medical problem that is also a societal problem,” Dr. Brady said. “We're trying to tackle both at the same time and want to jump quickly to interventions.”
As a positive example, she noted that providers can improve the institutions that serve the underserved.
“We'd like to have a simple solution that works across medically underserved and minority populations,” Dr. Brady said. “What we end up having is a more complicated problem. With real resources dedicated to ending disparity, we can make progress.”