We used data from the 2013/2014 to 2017/2018 National Health and Nutrition Examination Survey (NHANES) (n = 28,061), as the tobacco questions were consistent during this time period.17 NHANES is a nationwide probability sample of the US civilian noninstitutionalized population conducted continuously and released in 2-year cycles.18
We included former smokers who were eligible for LCS (n = 532) (Fig. 1). Eligible former smokers who were missing cotinine (n = 33), using nicotine replacement therapy (n = 12), or had previously been diagnosed with lung cancer (n = 20) were excluded, for a final sample size of 472. This analysis of deidentified, secondary data was exempt from review as compliant with the policy of the University of California, Davis Institutional Review Board.
Former smokers were classified as those who had smoked at least 100 cigarettes in their life and responded, “Not at all” to the question “Do you now smoke cigarettes?”. Former smokers were eligible for LCS if they were 50–80 years old, had a 20 pack-year smoking history, and quit within the past 15 years. Pack-years were calculated by multiplying the calculated cigarette packs (number of cigarettes smoked per day divided by 20 cigarettes per pack) by the number of years smoked. Years since quitting was self-reported based on the following question “How long has it been since you quit smoking cigarettes?”. Years since quitting smoking were also categorized at 0–3, 4–6, and > 6 years, based on sample sizes. Age was top coded at 80, and therefore included adults 80 and older (n = 39) as eligible for LCS.
Study Outcomes: Tobacco Use and Exposure
This study included two outcomes: recent tobacco use and exposure. Former smokers were classified as having recent tobacco use based on two sources: (1) self-reported use of any tobacco products in the past 5 days or (2) cotinine levels above the cut point for each racial/ethnic group (≥ 5.92 ng/mL (non-Hispanic Black), 4.85 ng/mL (non-Hispanic White), 0.84 ng/mL (Hispanic; this level reflects the largest subgroup being Mexican Americans19), and 3.08 ng/mL (all other).20 Former smokers without recent tobacco use were classified as having recent tobacco exposure if their cotinine levels were greater than 0.05 ng/mL. Although NHANES now has a lower limit of detection (0.011 ng/mL), we used 0.05 ng/mL for historical comparison with the general population.
Former smokers who recently used a tobacco product were further classified into using cigarettes, e-cigarettes, or another tobacco product (pipes, cigars, and smokeless tobacco were categorized together due to small sample sizes). We did not consider dual use of tobacco products due to small sample sizes. Former smokers were classified as living with a household smoker, if they responded with one or more to “How many people who live here smoke cigarettes, cigars, little cigars, pipes, water pipes, hookah, or any other tobacco product?” Former smokers were classified as being recently exposed to indoor smoke outside the home, if they reported that they spent time in an area (work, restaurant, bar, car, another home, or other indoor area) with someone else smoking in the past 7 days.
Demographics and Medical Conditions
Self-reported demographic characteristics included age (50–64, 65–74, and 75–80 years), race/ethnicity (Hispanic, non-Hispanic White, non-Hispanic Black, and non-Hispanic Asian), gender (male, female), income (less than or equal to 100% of the federal poverty level), education (less than a high school graduate or GED and some college or a college graduate), married or living with a partner, type of health insurance (private, public [Medicare, Medicaid, or other], or uninsured), and survey cycle. Self-reported medical conditions included respiratory disease (emphysema, chronic bronchitis, or asthma), coronary heart disease or stroke, cancer, and diabetes. Depression was categorized as yes (mild, moderate, severe) or no based on the Patient Health Questionnaire (PHQ-9), which is a self-reported assessment based on DSM-IV signs and symptoms for depression.
Prevalence estimates of recent tobacco use and recent tobacco exposure among former smokers were estimated for each demographic characteristic, tobacco-related behavior, and medical condition. Differences in prevalence estimates were assessed using the chi-square test. Adjusted logistic regression analysis was used to examine the association between the characteristics and each outcome, adjusted for each characteristic (except for previous cancer). Sensitivity and specificity were calculated for self-reported recent tobacco use and cotinine levels with the racial/ethnic cut points for tobacco use.20 The sample size for the sensitivity and specificity analysis was 450 because 22 people were missing data on self-reported tobacco use in the past 5 days. To assess the stability of each prevalence estimate, relative standard errors (RSE) (standard error divided by estimate) were calculated. All analyses were weighted to account for differential sampling probabilities and response rates, and standard errors were adjusted for the survey design using survey-specific procedures in SAS 9.4 (SAS Institute, Cary NC).