Unpacking differences among estimates of cigarette smoking prevalence in US adults as measured by the Behavioral Risk Factor Surveillance System (BRFSS), National Health Interview Survey (NHIS) and National Survey on Drug use and Health (NSDUH) from 2002-2011

To better understand differences in estimates of current cigarette smoking (current-CS) in US adults obtained from national surveys, we compared trends in daily and nondaily cigarette smoking from BRFSS, NHIS, and NSDUH. Current-CS is defined by NSDUH as any cigarette use during the previous 30 days and by NHIS and BRFSS as currently smoking daily or some days among persons who had ever smoked >=100 lifetime cigarettes. NSDUH defines daily-CS as smoking on all 30 and nondaily-CS as smoking on 1-29 of the previous 30 days. NHIS and NSDUH are household surveys. In 2011, the BRFSS telephone survey began weighting its data with a raking methodology and including data from cell-phone users. NSDUH estimates of current-CS were 4-5 percentage points higher than those of NHIS during 2002-2011 and 5-7 percentage points higher than those of BRFSS during 2002-2010. NSDUH estimates of daily-CS decreased over time and were virtually identical to NHIS while BRFSS was substantially lower (e.g., 2010: NSDUH and NHIS 15.1% vs. BRFSS 12.2%). However, 2011 estimates of daily-CS were not substantially different (NSDUH-14.7%; NHIS-14.8%; BRFSS-14.3%). NSDUH estimates of nondaily-CS were substantially higher than those of NHIS and BRFSS (9-10% vs. 4-6%) from 2002-2011. In applying the 100 cigarette criterion of NHIS and BRFSS to NSDUH analyses, trends in daily smoking were not affected, but estimates of nondaily-CS became lower, especially among 18-24 year olds. Differences in nondaily-CS, the 100 cigarette criterion, and BRFSS methodology explain some of the between-survey differences. Additional potential explanatory factors will be discussed.