JMIR Publications

ABSTRACT

Background:

Electronic cigarette (e-cigarette) device and liquid characteristics are highly customizable; these characteristics impact nicotine delivery and exposure to toxic constituents. It is critical that we understand the optimal methods for measuring these characteristics to accurately assess their impacts on user behavior and health.

Objective:

In order to inform future survey development, we assessed the agreement between responses from survey participants (self-reports) and participant uploaded photos as well as the quantity of useable data derived from each approach.

Methods:

Adult regular e-cigarette users (5+ days/week) aged 21+ (n=1209) were asked questions about and submitted photos of their most used e-cigarette device (n=1209) and liquid (n=1132). Device variables assessed included brand, model, disposable/reusable, device refillable/non-refillable, device display, and adjustable wattage/voltage. Liquid variables included brand, flavor, nicotine concentration, nicotine formulation, and container size. For each variable, percent agreement was calculated where self-report and photo data were available. Krippendorf’s alpha and intra-class correlation (ICC) were also calculated for categorical and continuous variables, respectively. All results were stratified by device (disposable, reusable with disposable pods/cartridges, reusable with refillable pods/cartridges/tanks) and liquid (custom, non-custom) type. The sample size for each calculation ranged from 47 (device model of disposables) to 1150 (device disposable/reusable).

Results:

Percent agreement between photos and self-reports was substantial to very high across device and liquid types for all variables except nicotine concentration. These results are in line with Krippendorf’s alpha calculations, except where prevalence bias was suspected to influence results. ICC results for nicotine concentration and container size were lower than percent agreement, likely because ICC accounts for the level of disagreement between values. Agreement also varied by device and liquid type. For example, percent agreement for device brand was higher among users of reusable devices (94%) than disposable devices (75%). Disagreement between photos and self-reports may be the result of low participant knowledge of characteristics, user modifications of devices inconsistent with the manufacturer intended use, inaccurate/incomplete information on websites, or submission of photos that are not a participant’s most used device/liquid. Additionally, the number of excluded values (e.g. self-report was ‘don’t know’, no photo submitted) differed between self-reports and photos; for questions asked to participants, self-reports had more useable data than photos for all variables except device model and nicotine formulation.

Conclusions:

Photos and self-reports yield data of similar accuracy for most variables assessed here: device brand, device model, disposable/reusable, adjustable wattage/voltage, display, refillable/non-refillable, liquid brand, flavor, and container size. Self-reports provided more data for all variables except device model and nicotine formulation, for which photos provided more data. Using these two approaches simultaneously may optimize data quantity and quality. Future research should examine how to assess nicotine concentration and variables not included here (e.g. wattage, resistance) and the resource requirements of these approaches.