Background and challenges to implementation:
Accurate estimation of smoking prevalence is important in a smoking cessation policy. Previous study using the Korean National Health and Nutrition Examination Survey (KNHANES) data compared women's self-report and urinary cotinine, and found that smoking prevalence is underestimated by self-report suggesting smoking concealment by female smokers due to stigma and cultural influence. We aimed to investigate whether female smokers report their smoking status consistently and related it to their smoking patterns (daily vs. intermittent).

Intervention or response:
A cross-sectional, online panel survey was conducted with 2,080 women aged 19 to 39. We first administered standard smoking status questions of KNHANES, and positioned unrelated health questions in between, and repeatedly asked about their smoking status and patterns. Then we cross-tabulated them to determine consistency of their self-report.

Results and lessons learnt:
By initial self-reports, 366 (17.6%) were current smoker, 120 (5.8%) were former smoker, and 1594 (76.6%) were never smoker. At second question, 102 (6.3%) and 223 (14.0%) women initially declared themselves as non-smoker answered that they are current smoker and former smoker. Current and former smokers at first smokers consistently reported their status at second questions. Among 256 daily smoker at second questions, 251 (98.0%) reported they are smoker at first question. Among 210 intermittent smokers, only 115 (54.8%) reported they are smokers at first questions.

Conclusions and key recommendations:
At least 22% of current smokers (102/468) reported they are not smoking at first standard KNHANES smoking status question. Most of them were intermittent smoker. Survey methods to reduce underestimation of female smoking prevalence are needed.