print   Bookmark and Share  |  RSS  |  A  A  A

Health Disparities Initiative  Funding and Activities

The Health Disparities Initiative leads multiple funded grants, projects, and other activities related to cancer health disparities in Georgia, the Southeast, and in the United States. A description of each activity is included herein.

Summary of Initial Grady-based Projects Within the Winship Cancer Institute Health Disparities Initiative

Assessing Treatment Outcomes for Patients with Non-Hodgkin’s Lymphoma

PI:  Christopher Flowers, MD

Non-Hodgkin’s Lymphomas (NHLs) are a heterogeneous group of cancers that are characterized by abnormal growth of tissue in the lymphatic system. It is the fifth most common cancer in both men and women, and the incidence of lymphoma continues to increase among all age groups in Georgia and in the nation. In 2003, 53,400 new cases of non-Hodgkin’s lymphomas (NHL) were diagnosed in the United States. The lymphatic system comprises of the tissues, organs and vessels that produce, stores and deliver cells that fight infection, or ‘lymphocytes’. Of these, there are two main classes: T and B lymphocytes. ‘T-cells’ are responsible both for cell-mediated immunity and for stimulating ‘B-cells’. When activated, B-cells produce antibody. Lymphoma may be classified as a B-cell or T-cell NHL, depending on whether it is B or T lymphocytes that are proliferating at an abnormal rate.

Approximately 85% of all NHLs are of B-cell origin and the remaining 15% of T-cell origin (2). Diffuse Large B Cell Lymphomas (DLBCLs) are a pathological subclass of NHLs grouped together for clinical purposes (3). All are composed of large cells with vesicular nuclei, prominent nucleoli, basophilic cytoplasm and a moderate to high proliferation fraction. They typically present as a nodal or extranodal mass with fast tumor growth associated with systemic symptoms, such as sweats, fatigue and fever. In about 40% of cases, these lymphomas appear in areas outside lymph nodes, including the digestive tract, skin, bone, thyroid and testes.  Surgery is typically carried out for diagnostic purposes and, once the DLBCL is identified, it is staged to find out how far the disease has spread. The standard staging system for non-Hodgkin’s lymphomas is determined by both the number of sites of involvement and the presence of disease above or below the diaphragm (4).

An individual’s risk of developing non-Hodgkin’s lymphoma during his lifetime is about 1 in 50. The risk of dying from NHL during one’s lifetime is about 1 in 100. Overall survival statistics for NHL typically are not helpful because survival depends on the type of lymphoma. The Revised European-American Lymphoma/World Health Organization (REAL/WHO) classification schema for non-Hodgkin’s lymphomas was devised to help aid in prognosis and treatment decision making.  Since cause for increasing incidence of NHL is unknown, new approaches are needed to increase the cure rates for NHL. Current treatments include conventional chemotherapy, radioimmunotherapy, and stem cell transplants as well as watching and waiting to start treatment (5-6).  Many prognostic factors, such as age and number of relapses, can influence how a patient will respond to certain treatments (7).  The goal of this proposal is to determine which factors influence disease free survival and overall survival in the treatment of relapsed non-Hodgkin’s lymphomas. 

The objectives of this study are:

  1. To augment data already obtained through Emory Healthcare databases and medical record reviews with additional data from detailed chart reviews at Grady Memorial Hospital to identify a large, diverse sample of patients with a diagnosis of non-Hodgkin’s lymphoma.
  2. To assess outcomes of relapsed non-Hodgkin’s lymphoma with stem cell transplantation and chemotherapy as treatments.
  3. To find the best way to identify patients with a particular non-Hodgkin’s lymphoma diagnosis within databases.
  4. To examine treatment disparities among patients diagnosed with non-Hodgkin’s lymphoma.
  5. To identify various patterns of care after diagnosis.

In essence, the Grady based sample will serve as a comparator population to our existing Emory cohort.  The initial focus would be on patients diagnosed with Diffuse large B-cell lymphoma (DLBCL), where we have identified 499 Emory cases from 1980-2008. With the Grady population we have identified ~285 DLBCL cases from 1984-2009, with an African American/Caucasian ratio of 74%/26%.  We hope to eventually collect data on about 1250 (~80%AA/20%Caucasian) identified Grady patients with lymphoid cancers (lymphoma/myeloma).  Reaching this goal is going to be dependent on the availability of the medical records. 

Back to top

Determinants of Patient Dropout from Cancer Treatment & Followup

PI:  Joseph Lipscomb, PhD

Previous studies have shown that cancer patients often do not receive care that matches what is considered standard or optimal based on clinical guidelines. However, the actual percentages of cancer patients who do not complete their prescribed therapy, or the reasons why, are not known. Neither is it known if the current data collection systems used by tumor registries and hospitals or other health care systems are adequate to identify or predict which patients may “drop out” of their treatment plan. Thus, the purposes of this study are: 1) to look at how often patients with breast, colorectal, lung and prostate cancer do not complete their prescribed treatment during the first year after a cancer diagnosis; 2) to investigate factors related to the cancer, the patient, the healthcare provider, and the healthcare system that are associated with patient dropout; and 3) to determine which sources of data are needed to answer the questions raised in this proposal. For this study, data from patients treated at Grady Hospital's Georgia Cancer Center of Excellence, will be reviewed. Data from the Georgia State Cancer Registry will be used to identify all cases of breast, colorectal, lung and prostate cancer diagnosed between 2004-2007. Additional data sources (e.g. medical record) will be used to determine which of those patients did not complete treatment as prescribed. It is estimated that about 290 breast, 185 colorectal cancer cases will be reviewed for this data analysis. Trained data abstractors will obtain data from each case’s hospital and/or clinic medical records, in addition to cancer registry data, in order to evaluate specific information related to the patient’s treatment. These abstracted data will be entered into an electronic format for further analysis. Such data will then be used to answer the questions that pertain to the study goals and determine patient dropout rates, possible reasons for early discontinuation of therapy, whether factors can be identified that might predict patient dropout from prescribed cancer treatment, and whether the data sources used are adequate to answer these types of questions about patient care. 

This study is building on an earlier CDC-supported study based in the 33-county, largely rural area of Southwest Georgia.  We will be using a modified version of the data collection instrument that had been created for that study.  Results from the Grady-based study will be compared with the emerging findings from the Southwest Georgia study. 

Back to top

Assessment of National Quality Forum Metrics in Colorectal Cancer in the Grady Hospital Population

PI:  Joseph Lipscomb, PhD

The Commission on Cancer (CoC) of the American College of Surgeons (ACoS) submitted quality of care measures for breast and colorectal cancer to the National Quality Forum (NQF) in response to its call for proposed breast measures in late 2004 and colorectal measures in early 2005.  Measures were reviewed by the CoC’s breast and colorectal disease site teams prior to their submission to the NQF for consideration.  A NQF Steering Committee for quality of cancer care measures was charge with assuring that pertinent stakeholders had appropriate opportunity to review and provide input on the measures under consideration.  Two Technical Panels assembled by the NQF made up of breast and colorectal experts in the areas of surgery, radiotherapy, medical oncology, health care consumers, and health services research provided technical evaluation of the proposed measures.  The NQF Steering Committee and Technical panels reviewed measures using four criteria:

  1. importance: the extent to which a measure reflects variation that has the potential for improvement;
  2. scientific acceptability: that a measure is reliable, valid, precise, and adaptable to  patient preference;
  3. usability: information produced as part of the measure could be used to make decisions and/or take actions, and that reported performance levels were statistically, and clinically meaningful:
  4. feasibility: that data can be obtained within the normal flow of clinical care and that implementation of the measure was achievable.

The specific aim of this study is to evaluate the measures of breast and colorectal cancer quality as identified by the National Quality Forum (NQF) for patients diagnosed at Grady Memorial Hospital.  The focus will be on the estimated 250 Stage I-III breast cancer cases and roughly 185 colon and rectal cases diagnosed at Grady over the 2004-2007 period.

The data for NQF measures assessment are being collected concurrently on the subset of patients in the determinants-of-dropout study sample who also meet NQF inclusion criteria.  For this purpose, the data collection instrument originally developed for the Southwest Georgia study was modified to capture those instances when chart documentation indicates the patient was offered or had recommended to her/him an intervention that was NQF-compliant, whether or not the intervention was recorded as having been received.  For example, the NQF recommends that hormone-receptor negative women diagnosed with non-metastatic breast cancer (and meeting other inclusion criteria) “receive or be offered” combination chemotherapy.  Hence, a full assessment of NQF adherence requires data on both utilization and any communications bearing on the decision not to receive therapy.  A direct and significant byproduct here is information on the extent to which reasons for non-treatment are documented in circumstances (such as with these NQF measures) where the general recommendation is to treat.

Back to top

Enhancing Treatment Decision-Making: Helping Underserved Populations Understand Risk Communications Related to Predictive Biomarkers

PI:  Theresa Gillespie, PhD

Advances in knowledge of and emerging technologies related to genomics and biomarkers in breast cancer have the potential to affect existing standards and quality of care and important outcomes such as progression, mortality, survival, and quality of life.  The use of gene expression biomarkers as prognostic and therapeutic indicators may serve as a key component of decision-making (DM) in cancer. One such technological application is the 21-gene assay used to predict distant disease recurrence in lymph node-negative (LN-), estrogen-receptor positive (ER+) breast cancer (Oncotype DX™, Genomic Health, Redwood, CA). Clinical trials have demonstrated that about 85% of women with LN-, ER+ early stage breast cancer (ESBC) may be overtreated if given cytotoxic chemotherapy in addition to hormonal therapy after definitive surgery (Paik 2004). Due to the significant morbidity and costs associated with systemic chemotherapy, efforts have been made to better define risk of recurrence.  The Oncotype DX™ assay assigns a Recurrence Score (RS) to indicate low, intermediate, or high risk of disease recurrence and has now become standard of care in ESBC (Oratz et al 2007). Ongoing studies of gene expression profiling have shown the likely expansion of the Oncotype DX™ assay to other tumor stages, or development of similar assays for other clinical scenarios.  Thus, the implications of the use of a predictive assay to aid DM for ESBC patients are profound.  However, previously-published research has shown the difficulty inherent in patient understanding of risk calculations generated by new technologies (Grimes 1999; Fuller 2001), particularly for patients with low literacy overall and low health literacy in particular(Guerra 2005). Risk communication is an essential aspect of patient DM, since most cancer screening and treatment options are associated with some risk:benefit ratio and varying degrees of uncertainty (Gattallari 2002). Understanding risks and their quantitative expression as probabilities (termed “numeracy”) is considered key to informed DM in situations where such risks are pivotal in deciding between different screening or treatment options, such as in breast cancer. This study will investigate patients’ understanding of numeracy as represented by risks of diagnosis or recurrence as predicted by genomic biomarkers. These findings will be applied to pilot test an intervention to improve patient understanding of risk communications in associated with predictive biomarkers in breast cancer.

Treatment decision-making (DM) is a critical health management task faced by cancer survivors at time of initial diagnosis, relapse or recurrence; for palliative care; and as related to symptom management. Much of cancer decision-making is dependent on understanding risks, including risks of adverse events, risks of recurrence of disease, or risk of death, based on what options for care may be chosen. This exploratory proposal will evaluate comprehension of risk communications related to biomarkers in breast cancer by 170 patients seen for screening or breast cancer care in the Avon Foundation Comprehensive Breast Center (AFCBC) at the Grady Cancer Center of Excellence. This study will provide initial feasibility and acceptability testing of an intervention designed to enhance comprehension of treatment options and thus facilitate informed treatment DM.

Back to top