A Population-based Smartphone Survey on Tobacco Use

Smartphone web access is rapid, convenient, and confidential, features that support a compelling rationale for using smartphone technology to engage respondents in ongoing public health surveillance.  The purpose of this feasibility study was to assess the use of smartphones as a mode for administering a population-based web survey about tobacco use behaviors. Starting with a national sampling frame of 25,000 random-digit-dial cell phone numbers, we successfully screened a sample of 1,446 respondents aged 18 to 65 years for eligibility to participate in the study.  1,068 respondents were identified as eligible as smartphone users, of whom 528 (49%) agreed to participate in the study.   Study participants were texted links to complete two brief web surveys over the course of two weeks. Smartphone survey data were compared to National Adult Tobacco Survey (NATS) landline and cell phone data. Specifically, we 1) examined demographic characteristics of respondents were examined using pairwise contingency tables; 2) examined the bivariate relationship between survey modes and tobacco use behaviors; and 3) developed logistic regression models for each type of tobacco use, adjusting for age and race/ethnicity,  to examine whether survey mode affected responses. Smartphone respondents reported significantly lower odds of using cigarettes, cigars, e-cigarettes and water pipes than those in the NATS cell phone survey, but the only difference observed between the smartphone and NATS landline surveys was reported use of cigars. The study findings provide insights into the feasibility of using smartphones in public health surveillance. We conclude that it is possible to conduct a population-based tobacco-related interview with respondents by smartphone, however, additional studies are needed to ascertain the validity and reliability of outcome data obtained using smartphone surveys to assess tobacco use behaviors.