Search of CEDR Data File Sets

Description:

This analytic data file set consists of one file generated for the lung cancer case-control study of workers at the Portsmouth Naval Shipyard (PNS) in Kittery, Maine. This study was performed by James Yiin et al., National Institute for Occupational Safety and Health (NIOSH).

Workers at the Portsmouth Naval Shipyard (PNS) in Kittery, Maine have been subjects of a number of epidemiologic investigations with different results, particularly for lung cancer and leukemia mortality. The purpose of this nested case-control study was to evaluate the relation between lung cancer risk and external ionizing radiation exposure while adjusting for potential confounders that included gender, radiation monitoring status, smoking habit surrogates (socioeconomic status and birth cohort), welding fumes and asbestos.

Cases and controls were from a cohort mortality study of 37,853 male and female civilian workers of all races ever employed at the shipyard between January 1, 1952 and December 31, 1992, with vital status updated through 1996. Vital status ascertainment for study subjects was updated by matching a roster of study subjects who had not already been identified as deceased with Social Security Administration (SSA) Master Death Tapes and the National Center for Health Statistics' National Death Index (NDI) service. The roster was also submitted to the SSA for a presumed living search. Death certificates were requested from the states for individuals who were identified as deceased through sources other than NDI, which provides cause-of-death information for deaths occurring in 1979 or later. Deaths were then coded by a qualified nosologist to the revision of the International Classification of Diseases (ICD) in effect at the time of death. Pre-1979 death information was obtained through earlier submissions to the Social Security administration and numerous other sources. Cases (n=1,097) had an underlying cause of death classified as lung cancer. For each case, a risk set was developed comprising all workers who were at risk of dying from lung cancer and who lived to at least the age attained by the index case at the time of death. Using an incidence density sampling program, three controls were randomly selected from each risk set, for a total of 3,291 controls.

Details of the radiation exposures at PNS and the procedures for assembling and validating dosimetry data are described by Daniels and colleagues (2004), please see the references below. Recorded occupational doses resulted primarily from external exposures to low linear energy transfer penetrating radiation. Cumulative exposures (mSv) from all years monitored were calculated for each worker by summing the results from each monitoring cycle. A lag of 15 years was employed to discount any exposure that occurred 15 years prior to death or end of study. Radiation workers were reclassified as non-radiation workers if all their exposures were recorded in the lagged years. Non-radiation workers were assumed to be unexposed to ionizing radiation from occupational sources other than work-related medical X-ray examinations, since they were restricted from the radiological areas. Information describing work-related X-rays was abstracted from the medical records for each study subjects.

Employment histories included job titles, shop assignments, and dates of employment for study subjects found in PNS personnel records. Shops are occupational groups describing major shipyard tasks, while job titles describe specific functions. Exposure matrices for asbestos and welding fumes were created for these shop and job-title combinations from 1945 to 1996 in 3-year periods. Each exposure matrix cell's final estimate was the product of exposure intensity (in fibers/cm3 for asbestos or mg/m3 for welding fume) and frequency, divided by a protection factor, if a respirator was required, plus the product of background intensity and frequency of unexposed days (background adjustment). Cumulative exposures for asbestos (fiber-days/cm3) and welding fume (mg-days/m3) were calculated for each study subject by summing all matching shop/job/time combinations in their employment histories. A lag of 15 yeas was also used for chemical exposures.

We conducted conditional logistic regression of lung cancer risk and external ionizing radiation in three model forms: log-linear (odds ratios at 10 mSv), linear excess relative risk (ERR per 10 mSv) and categorical (1-<10, 10-<50, and >50 mSv versus baseline 0-<1 mSv). Potential confounding variables modeled explicitly were assumed to have a log-linear relationship with cancer risk, and included: sex (female versus male), radiation monitoring status (yes versus no), socioeconomic status (skilled labor/unknown and administrative support/unskilled labor versus professional/technical, i.e., assumed smoking categories: intermediate and higher versus lower), birth cohort, welding fume (at 1,000 mg-days/m3) and asbestos (at 1,000 fiber-days/cm3). Categorized welding fume and asbestos were also considered. For welding fume, TLV-1 equals to 1,200 mg-days/m3 (5 mg/m3 per day x 240 working days per year x 1 year) and for asbestos, TLV-10 equals to 240 fiber-days/cm3 (0.1 fiber/cm3 per day x 240 working days per year x 10 years). Odds ratios and the corresponding 95% profile likelihood-based confidence intervals were derived from parameter estimates and the associated standard errors of the regression model.

Demographic information as well as radiological and chemical exposures for all cases and controls are included in the PLUN2CEDR file. Demographics include a study-generated identifier for each subject, risk set identifier (each set contains 1 case and 3 matched controls), case/control indicator, gender, year of birth, birth cohort, socioeconomic status (surrogate for smoking status), and radiation monitoring status indicator. Exposure variables include cumulative occupational dose (mSv), categorized occupational dose, occupational dose plus work-related X-rays, cumulative exposures for asbestos (fiber-days/cm3) and welding fume (mg-days/m3), and categorized asbestos and welding fume exposures.

Number of data files: 1

Cohort Size: 4,388

Sexes: male and female.

Diseases: lung cancer.

Exposure Type: External exposures to occupational ionizing radiation, including work-related medical x-ray, plutonium and tritium. Asbestos and welding fumes.
Exposure Agent: Chemical agents and ionizing radiation
Method: External dosimetry. Film badge monitoring began at PNS on 1 July 1950 with Kodak Eastman. In 1957, this site began using a film badge incorporating the DuPont SX-233 film packet. This dosemeter design remain unchanged until July 1969 when the Kodak Type 3 Film Radiac Pack replaced it. TLD became the standard for personnel gamma exposure monitoring on 1 October 1974. Note: This information has been taken from the article Bias and Penetrating Photon Dose Measured by Film Dosemeters in an Epidemiological Study of US Nuclear Workers by R. D. Daniels and M. K. Schubauer-Berigan.

Sites: Portsmouth Naval Shipyard (PNS) in Kittery, Maine.

Contact Information:

Yiin, James H
Principal Investigator
Industrywide Studies Branch. Division of Surveillance, Hazard Evaluations, and Field Studies, National Institute for Occupational Safety and Health (NIOSH)
Ph.D., Senior Service Fellow
NIOSH, MS R-44, 4676 Columbia Parkway
Cincinnati, OH 45226
(513) 841-4271
Fax: (513) 841-4470
Email: JYiin@cdc.gov

CITATION:   CEDR Library ID: 2070 (for full bibliographic record)

CITATION:   CEDR Library ID: 1996 (for full bibliographic record)

CITATION:   CEDR Library ID: 2003 (for full bibliographic record)

CITATION:   CEDR Library ID: 3972 (for full bibliographic record)

CITATION:   CEDR Library ID: 2356 (for full bibliographic record)

ADDED REFERENCE:   CEDR Library ID: 2054 (for full bibliographic record)

ADDED REFERENCE:   CEDR Library ID: 2682 (for full bibliographic record)