National Cancer Institute Home at the National Institutes of Health | www.cancer.gov

Health Disparities Interest Group September 2009 Seminar

Photograph of Dr. Bradley looking at a computer screen during her presentation.

Introduction

  • Dr. Nancy Breen called the meeting to order and asked participants to introduce themselves. She noted that this was the last meeting in the series arranged by Applied Research Program staff. Surveillance Research Program staff will arrange the next series of HDIG meetings.
  • Dr. Joan Warren introduced the guest speaker, Dr. Cathy J. Bradley, who presented "Inadequate Access to Surgeons: Reason for Disparate Cancer Care?"

Summary of Dr. Bradley's Presentation/Discussion

Background/Overview

  • Dr. Bradley's research is supported by NCI grant R01-CA101835-01, In-Depth Examination of Disparities in Cancer Outcomes. The grant's investigators include Dr. Bradley, of Virginia Commonwealth University (VCU), Charles W. Given of Michigan State University, Bassam Dahman of VCU, and Glenn Copeland of the Michigan Department of Community Health. It is important to include state colleagues when conducting research that involves data sets over which the researchers do not have custody. Also, it is difficult to find good data sets for studying health disparities, especially at the population level.
  • The Institute of Medicine's 2002 report, Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care, identified disparities in health care between whites, African Americans, and Hispanics across various diseases, clinical factors, and clinical settings.
  • Disparities also have been found by insurance status, with poor outcomes associated with Medicaid insurance and being uninsured.
  • On a population basis, the best available sources for studying disparities are registry and claims data. These data include a large population, but no qualitative data and sparse patient outcome data.
  • In 1999, Dr. Bradley and others began to link Michigan Medicaid and Medicare claim files to the Michigan Tumor Registry for the years 1996 through 2000. These data belonged to the state, making it necessary for the researchers to develop cooperative relationships. This resulted in a statewide, population-based dataset that contained Detroit SEER Registry data. The researchers added data on physician specialty, Area Resource File data, and US Census data. Medicaid enrollment and claim files, Medicare Denominator and claim files, and tumor registry data also were added.
  • Despite having a collaborative team of multidisciplinary academics and state employees, federal funding support, knowledgeable and cooperative personnel, and high-quality tumor registry and Medicaid data, it took more than 2 years to assemble the data. It is crucial to build sufficient time for data compilation into project timelines.
  • Obtaining the data involved identifying various data sources and custodians, negotiating different chains of command, and obtaining required approvals (Institutional Review Boards, HIPAA waiver, Data Use Agreements).
  • The researchers linked the tumor registry, Medicare, and Medicaid data sets using probabilistic, deterministic, and prescribed matching and other techniques.
  • Photograph of Dr. Bradley standing at the head of a long table, with the other HDIG attendees seated around it.
    Data characteristics:
    • Strengths:
      • Statewide with nearly comprehensive claims for the elderly.
      • Spanned several years.
      • Included Medicaid enrollment history.
    • Limitations:
      • Absence of qualitative data.
      • Missing key variables regarding physicians (e.g., specialty) and patients (e.g., caregiver status, location).
  • Key points to consider when conducting disparity studies that involve the dually eligible:
    • Many dually eligible individuals are nursing home residents and, as a result, are not ideal candidates for cancer care. Many studies exclude these individuals.
    • Medicaid enrollment can be highly variable. An individual's history with regard to Medicaid enrollment can affect study outcomes. For example:
      • Some individuals are ever dually eligible.
      • Some are continuously enrolled 12 months prior to diagnosis.
      • Some are enrolled during the month before diagnosis.
      • Some are enrolled during the month of diagnosis or within 3 months following diagnosis.
    • Medicaid patients receive "comprehensive care," which includes coverage for prescription drugs, Medicare co-pays and deductibles, nursing home care, and other services not covered by Medicare. Medicaid reimburses for these services at very low rates; many providers do not consider it worthwhile to submit a claim for such services.
      • Dr. Bradley and colleagues found that Medicaid enrollment files were more useful than the actual claim files.

Research

  • The purpose of this research study was to examine the differences in cancer diagnosis, treatment, and outcomes between dually eligible and Medicare "only" patients; to attempt to determine if there are disparities and quantify them; and to attempt to determine why disparities exist.
  • Findings regarding dually eligible cancer patients:
    • Late stage at diagnosis (many are not staged and/or are diagnosed at death).
    • Nearly 60 percent enroll in Medicaid following a cancer diagnosis.
    • Prior health care and insurance status are uncertain; many have comorbid conditions.
    • Less likely to have chemotherapy and/or surgery.
    • Poor survival.
  • Findings regarding dually eligible colon cancer patients:
    • Relative to Medicare patients, less likely to receive chemotherapy.
    • Less likely to be evaluated by an oncologist (though it is unknown whether they are less likely to be referred to an oncologist).
    • (Reference: Bradley CJ, Given CW, Dahman B, Fitzgerald TL. Adjuvant chemotherapy after resection in elderly Medicare and Medicaid patients with colon cancer. Arch Intern Med 2008 Mar 10;168(5):521-9.)
  • Findings regarding dually eligible lung cancer patients:
    • Half as likely as Medicare "only" patients to undergo a resection.
    • More likely to receive radiation than Medicare "only" patients.
    • (Reference: Bradley CJ, Dahman B, Given CW. Treatment and survival differences in older Medicare patients with lung cancer as compared with those who are dually eligible for Medicare and Medicaid. J Clin Oncol 2008 Nov 1;26(31):5067-73. Epub 2008 Sep 15.)
  • The researchers found that the claim data did not include patients treated at small, rural hospitals without a cancer registry.
  • Preliminary findings:
    • Dually eligible patients were less likely than Medicare patients to be evaluated by a surgeon.
    • Approximately one-half (52%) of non-small cell lung cancer dually eligible patients were evaluated by a surgeon, whereas nearly three-fourths (74%) of Medicare patients were evaluated by a surgeon.
    • When patients were evaluated by a surgeon, the likelihood of having a resection was similar.
  • Photograph of Dr. Bradley gesturing toward a diagram on the projector screen.
    Conclusion: The key to equivalent care appears to be related to access to specialists such as surgeons who have experience in performing lung or colon resections. (Reference: Bradley CJ, Dahman B, Given CW. Inadequate access to surgeons: reason for disparate cancer care? Med Care 2009 Jul;47(7):758-64.)
  • Implications for practice:
    • Determine whether poor follow-up on the part of dually eligible patients is related to refusals, cancellations, no-shows.
    • Determine the barriers to seeing dually eligible or other low-income patients.
  • Implications for policy:
    • Identify specialists who are accepting dually eligible patients and why.
    • Understand issues related to access.
      • Specialists may accept a limited number of dually eligible patients per day.
      • The number of specialists available to dually eligible patients may be inadequate.
  • Implications for research:
    • Examine more recent, nationwide data.
    • Examine facility characteristics.
    • Conduct qualitative research on patients regarding referrals and followups.
    • Compare to patients in managed care organizations.
  • "Bigger picture" questions:
    • Need better understanding of why disparities in care exist.
      • Difference in referrals, refusals, access.
      • Difference in health status.
    • Is less care appropriate?
      • Differential care is not appropriate, but what is the appropriate level of care for a given clinical profile?
    • Are there alternative treatments that provide good quality-of-life outcomes?
      • Are patients being supported in other ways?

Discussion

  • Patients followed by a registry tend to have more complete data. Lack of complete data is not necessarily reflective of quality of care.

HD*Calc Announcement

  • The Health Disparities Calculator (HD*Calc), a statistical software designed to generate multiple summary measures to evaluate and monitor health disparities, was released on September 23, 2009.
  • Participants interested in being trained on the use of HD*Calc in November 2009 should contact Annie Sampson at sampsona@mail.nih.gov.

Next HDIG Meeting

  • SRP staff will circulate information about the next HDIG speaker presentation as it becomes available.

Last Modified: 03 Sep 2013