Smoking prevalence and association with sociodemographic variables in cancer clinical trial participants

Tobacco use (smoking) causes adverse clinical outcomes among patients with cancer, including increased cancer-related mortality. In participants in cancer clinical trials, the prevalence of tobacco use and the factors associated with tobacco use are not well described.

Data were examined from participants enrolled in SWOG cancer clinical treatment trials between 2016 and 2022 who reported their smoking status at trial enrollment. Baseline variables (smoking status, insurance type, zip code, and demographic factors) were obtained from patient registration forms. Bivariate and multivariable associations were examined via logistic regression.

Among 4326 patients enrolled in 29 trials, 48.1% reported currently/previously smoking, including 12.4% currently, 4.9% recently, and 30.7% formerly. Ever smoking was more commonly reported in males, patients aged ≥65 years, patients with Medicaid or no insurance, patients from areas of high socioeconomic deprivation, and rural patients. Patients of Hispanic ethnicity and Asian and Pacific Islander patients were less likely to have ever smoked. In multivariable regression, patients with lung cancer were most likely to report ever smoking compared to patients with breast cancer (odds ratio, 4.98; p < .001).

In the first comprehensive evaluation of smoking status among trial participants enrolled in National Cancer Institute network group treatment trials, nearly half reported ever smoking and one in six reported current or recent smoking. Smoking was more common among vulnerable population patients defined by demographic and socioeconomic factors. Tobacco use should be routinely assessed and reported in clinical trials to help reduce the negative cancer and overall health effects of persistent tobacco use and to address disparities among patients with cancer.