Methods of studying the relation of employment and long-term illness cohort analysis

Our plain-English summary of Methods of studying the relation of employment and long-term illness cohort analysis (1959).

The study

• Published: 1959
• Type of article: Journal article
• Our copy: from Johns-Manville's own medical library. The article is still under copyright, so we describe it here rather than post it.

In plain English

This 1959 paper from Ohio was not about asbestos. It described a method for long-term studies of workers' health using Social Security records held by the Bureau of Old-Age and Survivors Insurance (BOASI). To test the method, the authors followed male employees of one Ohio company, called "Company A", who worked there in 1937, through 1956. For many years the company had made beta-naphthylamine, which the article called a known cancer-causing agent, and benzidine, which it called a highly suspected one. Deaths were traced through BOASI records and, separately, through plant records, and were compared with Ohio death rates for 1950 to 1956. BOASI employment records were more complete than plant records. Among white males in the company groups, heart disease deaths were relatively low and cancer deaths relatively high compared with Ohio. Both sources showed an excess of bladder cancer deaths. Each source found 6 bladder cancer deaths. In the BOASI data, 4 of these were among white males, against 0.30 expected. The authors noted that both sources missed some deaths. The authors concluded that confirming a known occupational cancer hazard this way opened the way to further use of Social Security data.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Methods paper proposing cohort studies based on Social Security (BOASI) records, demonstrated with a historical cohort mortality study of one company's workforce. The company's workforce was traced both through BOASI records and through plant records.
• Who was studied: Male employees of one company ("Company A") located in a county in Ohio, employed in 1937 and followed through 1956. Deaths at ages 25 to 61 were compared with Ohio residents' death rates for 1950 to 1956. The number of employees is not legible in the scanned text.
• Exposure: Employment at a company that had for many years manufactured beta-naphthylamine, described as a known carcinogenic agent, and benzidine, described as a highly suspected carcinogen. Both chemical names are badly garbled in the scanned text. No asbestos exposure is studied.
• Main results:
  • BOASI (Social Security) records on employment were more complete than the company's plant records.
  • Among white males in the company cohorts, mortality from heart disease was relatively low and mortality from malignant neoplasms (cancers) relatively high compared with the state of Ohio.
  • An excess of bladder cancer deaths was found in the company cohort, detected in both the BOASI and the plant records. BOASI records showed 6 observed bladder cancer deaths in total, including 4 among white males against 0.30 expected. Plant records also showed 6 observed bladder cancer deaths. The other expected figures are not legible.
  • The authors described cancer of the pancreas as an example of a suggestive lead for other organs. This was not presented as a confirmed finding.
• The authors' conclusion: The authors concluded that the method confirmed a known occupational cancer hazard. In their view, this paved the way for further use of Social Security source data in studying working populations.
• Limitations the authors state:
  • Social Security records did not identify all deaths, because some survivors' benefit claims were not filed.
  • Not all 1937 employees could be traced through company data, so plant-record data did not capture all deaths or all bladder tumor cases.
  • Plant-record mortality data were relatively incomplete in the years before 1950.
  • The BOASI industry classification, even with a four-digit code, is still a broad category for study.

Browse return to medical research.