Premium
Comparing cancer risk estimates using occupational record linkage approaches in male Florida firefighters
Author(s) -
McClure Laura A.,
KoruSengul Tulay,
Hernandez Monique N.,
CabanMartinez Alberto J.,
Kobetz Erin N.,
Lee David J.
Publication year - 2021
Publication title -
american journal of industrial medicine
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.7
H-Index - 104
eISSN - 1097-0274
pISSN - 0271-3586
DOI - 10.1002/ajim.23205
Subject(s) - medicine , firefighting , confidence interval , odds ratio , logistic regression , cancer , occupational exposure , demography , occupational medicine , environmental health , chemistry , organic chemistry , sociology
Background Firefighters have an increased risk of cancer, but variations in reported results could be due to differences in occupational case ascertainment. This study compares cancer risk estimates generated by identifying firefighters from their occupational title available in the Florida Cancer Data System (FCDS) versus identification by a linkage method between the FCDS and the Florida State Fire Marshal's Office. Methods Florida firefighter employment records (1972–2012; n = 109,009) were linked with FCDS data (1981–2014; ~3.3 million records), identifying 3760 primary cancers in male firefighters. Using the FCDS occupational data field we identified 1831 male cancer cases in those classified as firefighters, first‐line supervisors of firefighting and prevention workers, fire inspectors, emergency medical technicians, or paramedics. Age and calendar year‐adjusted odds ratios (aOR) and 95% confidence intervals for firefighters versus non‐firefighters were calculated for both groups. Results For skin cancers the risk estimate for FCDS‐indentified firefighters was substantially lower than in the employment‐record‐linked firefighters (aOR = 1.06; 0.87–1.29 vs. 1.54; 1.37–1.73), but for endocrine system cancers it was greater (aOR = 2.36; 1.77–3.14 vs. 2.08; 1.71–2.53). Remaining cancer risk estimates were in the same direction for the two samples except for lymphoma (aOR = 1.10; 0.90–1.34 vs. 0.86; 0.75–0.99). Conclusion Reliance on occupational title in cancer registry records to characterize firefighter cancer risk may result in estimates that are over‐ or underestimated depending on cancer site. The authors recommend moving toward national linkages between cancer registries and certification or other administrative records, which are a vital resource for firefighter cancer research.