Impact of survey and weighting methods on tobacco use estimates

Audience: State tobacco control evaluators and epidemiologists who work on surveillance systems. Key Points: A major component to tobacco control programs is monitoring of key indicators, including prevalence of tobacco use. However, differences in estimates of current smoking prevalence among U.S. adults have been observed across several major surveillance systems including National Health Interview Survey, Behavioral Risk Factor Surveillance System (BRFSS), and National Adult Tobacco Survey (NATS). The purpose of this study is to identify potential methodological reasons for these differences. We will explore differences in sampling methods, sampling frames and coverage, data collection mode, survey topic, response rates and weighting methods across these surveillance systems. Data on current smoking and other tobacco indicators among U.S. adults will also be compared from two surveillance systems: BRFSS and NATS. Discussions will focus on differences in survey methodology and potential ways in which survey and weighting methods may be improved to produce more accurate estimates. Educational Experience: This presentation of survey methodology, weighting procedures, and data on smoking prevalence from various data systems will provide a better understanding of estimates from these dual-frame RDD surveys. Benefits: The audience will have the opportunity to understand changes in tobacco use estimates from national surveys resulting from recent changes in survey and weighting methods.