Ann Epidemiol. 2020 Nov 12:S1047-2797(20)30411-7. doi: 10.1016/j.annepidem.2020.11.004. Online ahead of print.
PURPOSE: In cancer patients, cigarette smoking is causally linked with increased mortality. We examined the relationship between smoking status at the time of diagnosis and cancer mortality to help estimate the scope of smoking cessation services required to meet the needs of cancer patients.
METHODS: We studied the ten most common cancers in Massachusetts, 2008-2013 including 175,489 incident cases and used smoking status at the time of diagnosis to provide smoking prevalence. We calculated adjusted hazard ratios of all-cause mortality comparing smoker subgroups.
RESULTS: Smoking prevalence was more than threefold higher for lung cancer and more than twofold higher for head and neck cancer and bladder cancer than in the general population. Cancer cases who smoked at the time of diagnosis had a higher adjusted mortality rate than cancer cases who were former smokers. The three sites with the highest increased hazard ratios comparing current smokers with former smokers were cancers of the thyroid (HR=1.67, 95%CI 1.14-2.45), head and neck (HR=1.65, 95%CI 1.39-1.95), and prostate (HR=1.60, 95%CI 1.36-1.90).
CONCLUSIONS: Smoking remains high among cancer patients. More widespread adoption of smoking cessation programs among cancer patients may play a substantial role in improving cancer morbidity and mortality.