Skip all navigation Skip to page navigation

DHHS Home | A-Z Site Map | Divisions | About Us | Contacts

NC Department of Health and Human Services
N.C. State Center for Health Statistics Home
N.C. Public Health Home
 
 

Posters Sessions

Central Cancer Registry

North Carolina Brain/CNS Cancer Mortality
and Associated Employment 1988-1997

By Allison Elledge, M.S.
North Carolina Central Cancer Registry
State Center for Health Statistics

for the 1999 Annual Meeting
North American Association of Central Cancer Registries

Table of Contents

  • Abstract
  • Adult brain cancer is a rapidly lethal disease, and there are complex and diverse theories regarding its etiology. In North Carolina there is an increased interest in the association between cancer and environmental exposures. In the years from 1980 to 1997, age-adjusted mortality rates of brain/other central nervous system (CNS) cancers have ranged from 3.0 to 4.9 deaths per 100,000 people (1970 U.S. standard population). During 1997 in North Carolina, 383 people ages 20 and older died from malignant brain and central nervous system neoplasms.

    Two issues were examined:

    1. Misclassification of brain cancer deaths, and
    2. Occupations/industries associated with brain cancer mortality in North Carolina.

    Since the brain is a frequent metastatic site, the issue of death certificate misclassification is important. It is possible that misclassification of brain/other CNS cancers on the death record may bias the results of an occupational study. A wide variety of white-collar and blue-collar industries has been associated with brain cancer mortality, but it has been very difficult to pinpoint particular environmental exposures.

    This study found:

    • 7.5% of the brain cancer deaths were misclassified.
    • Only 43.6% of the employment information on the Central Cancer Registry records agreed with the information on the matching death certificates.
    • A number of occupations and industries demonstrated increased risk of brain cancer mortality in North Carolina.
  • 1988-97 Brain/CNS Cancer Mortality Rates per 100,000 by North Carolina Economic Development Regions
  • Rates are Age-Adjusted to the 1990 US Census Population, Ages 20+

    Map of Rates by Economic Development Region
  • Percent Distribution of Employment by Sector, 1996 for North Carolina Economic Development Regions
    • Advantage-West, Piedmont Triad, Carolinas Partnership
      Percent Distribution of Employment by Sector, 1996 for North Carolina Economic Development Regions
      Results by Employment Sector - Part 1
    • Southeast, Research Triangle, Global Transpark, Northeast
      Percent Distribution of Employment by Sector, 1996 for North Carolina Economic Development Regions
      Results by Employment Sector - Part 2
  • PHASE I - Misclassification of Brain Cancer on the Death Certificate
    • Demographic profile of matches
    • Comparison of Occupations and Industries for matches

    The brain is a frequent metastatic site for other primary cancers and subject to misclassification on the death certificate. All patients diagnosed with cancer in 1995 were matched to all death records from January 1995 to May 1998. Results were recorded for people who were coded to ICD-9 191.0 - 192.9 on the death certificate in order to determine a misclassification rate.

    Results of Misclassification
  • Demographic Profile of 1995 Brain/CNS Cancer Cases Matches to Death Records from January 1995 to May 1998
  • Total Number of Matches Where Both the Registry Record and the Death Certificate Were Coded as Brain/CNS Cancer = 236

    Graphs of Demographic Profiles
  • Comparison of Occupation/ Industry Listed on 1995 Incident Record and Death Certificate
  • In order to have agreement, the CCR text must have named either the occupation or industry coded on the death certificate (1990 3-digit U.S. census occupation and industry codes):

    Number Percent
    Agreement 103 43.6%
    No Agreement* 133 56.4%
    Total 236 100.0%

    *Nine out of the 133 records with no agreement (7%) had "unemployed" or "unknown" on the death certificate. One hundred twenty-two (52%) of all Registry records were blank or had "unknown"/"retired"/"unemployed ."

     

    Occupations (census code) N Industries (census code) N
    Managers (003-017, 019) 16 Schools (842) 19
    Teachers (155-159) 9 Fabric Mills (142) 18
    Sales+ (258, 259) 6 Construction (060) 15
    Sales Supervisors (243) 6 Pulp Mills (160) 6
    Other Sales Workers (274) 5 Computer Equipment (322) 5
    Truck Drivers (804-805) 5 General Government (901) 5
       Hospitals (831) 5
    +Mining, manufacturing, wholesale Trucking Services (410) 5
  • PHASE II - Brain Cancer Mortality Risks by Occupation
  • Case Control Study of Employment
    Associated with Brain /CNS Cancer Deaths, by Race/Sex Group
    Ages 20 and Over

    Cases = Brain/ Other CNS Cancer Deaths from 1988-97 = 3,304
    Controls* = Non-Cancer Deaths from 1988-97 = 13,212

    *Controls were randomly selected from NC non-neoplasm death records (ICD-9 codes less than 140.0 and greater than 239.9) by age matching to brain/CNS cancer deaths, with a ratio of 4 controls to 1 case.

    Pie Charts of Cases and Controls

    Legend for Pie Charts

  • Statistically Significant Odds Ratios, White Males
  • Statistically Significant Odds Ratios for White Males

  • Statistically Significant Odds Ratios, Minority Males
  • Statistically Significant Odds Ratios for Minority Males

  • Statistically Significant Odds Ratios, White Females
  • Statistically Significant Odds Ratios for White Females

  • Statistically Significant Odds Ratios, Minority Females
  • Statistically Significant Odds Ratios for Minority Females

  • Conclusions
  • The main focus of this study was to look at brain cancer deaths and associated employment in North Carolina.

    Phase 1
    Phase 1 involved records matching.

    • Misclassification on Death Certificates
      Records of 1995 brain cancer patients from the N.C. Central Cancer Registry files were matched to death certificates in order to determine how many are being misclassified on the death certificate. It was suspected that since the brain is a site in the body where a primary cancer frequently spreads (metastasis), many death certificates therefore would be miscoded. Indeed, 7.5% of the death certificates were found to have been from patients whose first cancers arose elsewhere in the body.
      Information from Death Certificates coded as Brain Cancer should be used cautiously because of misclassification due to metastasis.
    • Missing Data
      It is optional for facilities submitting cancer patient records to the N.C. Central Cancer Registry to send information about work history (i.e., usual occupation), although many choose not to. Registry information on occupation, when available, compared favorably to the occupation and industry codes found on the death certificates.
      Registry records, however, are only marginally useful for occupation/industry-based risk or exposure studies, since data are frequently missing.

    Phase 2
    Phase 2 involved studying 10 years' worth of brain cancer death certificates to determine the occupations and industries most highly associated with brain cancer death. These were compared in a case-control fashion to employment information for non-cancer deaths by race and sex.

    • As in other studies, risk of brain cancer death was found in various white collar professions such as education, accounting, law, and dentistry.
    • Increased risk was also found for white males in professions associated with electronics and for minority males employed in automobile repair.
      Results should be interpreted with caution, since it is not known how long a person worked in the field represented on the death certificate. There are numerous professions that have been associated with the development of brain cancer, only a few of which have been proven conclusively.