Systematic Review and Meta-Analysis of Selected Cancers in Petroleum Refinery Workers

Our plain-English summary of Systematic Review and Meta-Analysis of Selected Cancers in Petroleum Refinery Workers, by A.R. Schnatter, M. Chen, E.A. DeVilbiss, et al (Journal of Occupational and Environmental Medicine, 2018).

The study

• Authors: A.R. Schnatter, M. Chen, E.A. DeVilbiss, R.J. Lewis, E.M. Gallagher
• Journal: Journal of Occupational and Environmental Medicine
• Published: 2018;60(7):e329-e342
• Type of article: Meta-analysis, Systematic review
• Cited by: 10 articles, as counted by NIH iCite
• Read it: PubMed · Publisher

Abstract

Objective: We studied the risk of 11 cancers of a priori interest in petroleum refinery workers.

Methods: Iterative searches identified 36 studies for the 11 cancer sites. Statistical heterogeneity and publication bias were assessed to enhance interpretation of meta-relative risks.

Results: Statistical heterogeneity was marked for mesothelioma, but was largely due to study quality. Higher quality studies showed a meta-relative risk (RR) of 3.22, (95% prediction interval 1.45 to 7.23). Melanoma (meta-RR = 1.23) and acute lymphoid leukemia (meta-RR = 1.51), showed results consistent with higher risk, but both were driven by one or two studies. Eight other cancer outcomes showed summary meta-RR's consistent with unity.

Conclusions: Most cancer outcomes are consistent with background risk in refinery workers. This work has clarified an excess mesothelioma risk, conditional on study quality stratification. Continued surveillance is warranted for melanoma and ALL.

In plain English

This systematic review and meta-analysis pooled 36 cohort-based studies of petroleum refinery workers to estimate the risk of 11 cancers compared with background rates, looking at refinery work in general rather than specific substances. Across 18 mesothelioma studies, the combined relative risk was 2.36, with wide variation (prediction interval 0.84 to 6.65). Studies using what the authors judged the best case-finding and comparison-rate methods gave 3.22 (prediction interval 1.44 to 7.23). Studies with comparison rates from mismatched time periods gave 1.10. Maintenance workers, whom the authors considered likely more often exposed to asbestos, showed 4.03 (95% CI 3.33 to 4.88). Excesses for acute lymphoid leukemia (1.51, 95% CI 1.08 to 2.13) and melanoma (1.23, 95% CI 1.09 to 1.38) were less certain, the first influenced by one large study, the second driven by incidence rather than mortality studies. For lung cancer (0.81, prediction interval 0.63 to 1.04), the authors judged risk not elevated and perhaps lower. The authors concluded mesothelioma showed a clear excess and that study methods, not differing asbestos exposure, more likely explained inconsistent earlier results.

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: Systematic review and meta-analysis of published and unpublished cohort-based studies (literature searched to March 2017)
• Who was studied: 36 cohort-based studies of petroleum refinery workers, mostly from the United States and Italy; study start years ranged from 1914 to 1983 and last follow-up from 1975 to 2007; most data applied to males
• Exposure: Employment in petroleum refineries in general, not specific substances. Asbestos is discussed in relation to mesothelioma and maintenance trades. No fiber types are named.
• Main results:
  • Of 18 studies reporting mesothelioma, 11 reported a significantly increased risk
  • All 18 mesothelioma studies combined: meta-RR 2.36 (95% CI 1.78 to 3.13), with marked heterogeneity, so the authors preferred the prediction interval of 0.84 to 6.65
  • Mesothelioma studies using both complete case ascertainment and concurrent background rates ('best methods'): meta-RR 3.22, prediction interval 1.44 to 7.23
  • Mesothelioma studies using non-concurrent background rates: meta-RR 1.10
  • Five studies of maintenance trades workers (welders, insulators, etc.): mesothelioma meta-RR 4.03 (95% CI 3.33 to 4.88)
  • Four studies with at least 20 years since first employment: mesothelioma meta-RR 3.45, prediction interval 1.75 to 6.82
  • Lung cancer in refinery workers: meta-RR 0.81 (random effects) and 0.82 (fixed effects), significantly low, but with significant heterogeneity; prediction interval 0.63 to 1.04. The authors judged the data most compatible with risk not being elevated and perhaps significantly lower than the general population.
  • Melanoma: meta-RR 1.23 (95% CI 1.09 to 1.38), driven mainly by two large incidence studies
  • Melanoma, mortality studies only: random meta-RR 1.06 (95% CI 0.85 to 1.31), not statistically significant; incidence rather than mortality drove the overall excess
  • Acute lymphoid leukemia: meta-RR 1.51 (95% CI 1.08 to 2.13), with one study weighted at 24.2%. The authors described the excesses for ALL and melanoma as less certain.
  • No clear signals for seven other cancers
• The authors' conclusion: The authors concluded that mesothelioma showed a clear excess in refinery workers. They also concluded that study methods, rather than differences in asbestos exposure, more likely explain why earlier studies did not consistently show elevated mesothelioma risk.
• Limitations the authors state:
  • Comparison with general-population rates (healthy worker effect) may bias summary risks low
  • Excesses in early, higher-exposure periods may have been missed because the most recent study updates were usually used
  • The analysis does not address risks from specific substances
  • For one study, regional rather than national rates were used for mesothelioma, which may underestimate the risk

Topics

The National Library of Medicine indexes this study under subjects our site covers: Refineries and petrochemical plants.

Medical terms in this study, explained in our Lexicon: Systematic Review, Meta-Analysis, Relative Risk, Mesothelioma.

Related studies

• Asbestos exposure and cancer mortality among petroleum refinery workers: a Poisson regression analysis of updated data — F. Montanaro, M. Ceppi, et al. (2004)
• Mortality and cancer morbidity in a cohort of Canadian petroleum workers — R.J. Lewis, A.R. Schnatter, et al. (2003)
• Long-term follow-up mortality study of petroleum refinery and chemical plant employees — S.P. Tsai, E.L. Gilstrap, et al. (1996)
• An updated mortality study of workers at a petroleum refinery in Beaumont, Texas — G.K. Raabe, K.W. Collingwood, et al. (1998)
• Kidney cancer and hydrocarbon exposures among petroleum refinery workers — C. Poole, N.A. Dreyer, et al. (1993)
• Asbestos-related cancers in refinery workers in the Australian petroleum industry — R. Gun, N.L. Pratt, et al. (2006)

Browse more research in Journal of Occupational and Environmental Medicine, or return to medical research.